Evolving Systems, Volume 7, Issue 4, December 2016

1. A speed FPGA hardware accelerator based FSBMA-VBSME used in H.264/AVC
Author(s): Amira Yahi, Salah Toumi, El-bay Bourennane & Kamel Messaoudi
Pages: 233-241

2. On cluster validity indices with its application to interleaved radar pulse separation through fuzzy-based representation
Author(s): A. H. Mazinan
Pages: 243-254

3. On fuzzy feature selection in designing fuzzy classifiers for high-dimensional data
Author(s): Eghbal G. Mansoori & Khadijeh S. Shafiee
Pages: 255-265

4. Adaptive PID controller design for wing rock suppression using self-recurrent wavelet neural network identifier
Author(s): Milad Malekzadeh, Jalil Sadati & Mojtaba Alizadeh
Pages: 267-275

5. Adaptive cow movement detection using evolving spiking neural network models
Author(s): Tao Gao & Nikola Kasabov
Pages: 277-285

6. An inventory model for controllable probabilistic deterioration rate under shortages
Author(s): Umakanta Mishra
Pages: 287-307

Review August - November 2016

 It's been a while since my last review post. Below are some of the interesting links I Tweeted about in the last few weeks.

1. 7 ways AI could be a good thing: https://www.theguardian.com/technology/2016/aug/07/seven-benefits-of-artificial-intelligence
2. A fairly detailed overview of IBM's Watson AI: http://www.techrepublic.com/article/ibm-watson-the-smart-persons-guide/
3. Using deep learning to automate spearphishing attacks: https://www.technologyreview.com/s/602109/this-ai-will-craft-tweets-that-youll-never-know-are-spam/
4. And publishers are still being dicks about Sci-Hub: https://www.insidehighered.com/news/2016/08/08/letter-publishers-group-adds-debate-over-sci-hub-and-librarians-who-study-it
5. Getting started understanding machine vision: http://www.kdnuggets.com/2016/08/seven-steps-understanding-computer-vision.html
6. AI applied to marketing and advertising: http://techemergence.com/artificial-intelligence-in-marketing-and-advertising-5-examples-of-real-traction/
7. The Nervana chip optimised for deep learning: http://www.nextbigfuture.com/2016/08/startup-nervana-making-deep-learning.html
8. AI and machine learning in finance: http://www.datasciencecentral.com/profiles/blogs/interview-with-flowcast-cto-ai-machine-learning-in-fintech
9. A guide to Google deepmind: http://www.techrepublic.com/article/google-deepmind-the-smart-persons-guide/
10. Tutorial on neural networks in R: http://www.kdnuggets.com/2016/08/begineers-guide-neural-networks-r.html
11. Brief step-by-step guide to building an expert system: http://www.techrepublic.com/article/40-year-old-ai-innovation-may-solve-your-big-data-problems/
12. Recent advances in quantum computers are promising for AI: https://www.datanami.com/2016/08/11/quantum-researchers-eye-ai-advances/
13. 3 thoughts from Yann LeCunn on why deep learning works so well: http://www.kdnuggets.com/2016/08/yann-lecun-3-thoughts-deep-learning.html
14. AI in healthcare-where it is, and where it's going: https://techcrunch.com/2016/08/12/the-healing-power-of-ai/
15. Why movies like The Terminator are not to blame for the bad journalism around AI: http://www.kdnuggets.com/2016/08/stop-blaming-terminator-for-bad-ai-journalism.html
16. Recognising hand gestures using IBM's neuromorphic chips: http://www.theverge.com/circuitbreaker/2016/8/12/12458330/samsung-ibm-truenorth-brain-chip-gesture-app
17. I've said it time and again-biased data produces biased models: http://www.nytimes.com/2016/06/26/opinion/sunday/artificial-intelligences-white-guy-problem.html?_r=2
18. 10 signs of a bad place to work: http://www3.forbes.com/leadership/ten-unmistakable-signs-of-a-bad-place-to-work/
19. The dangers of using biased data in training a model - with Donald Trump as the cautionary example: https://mathbabe.org/2016/08/11/donald-trump-is-like-a-biased-machine-learning-algorithm/
20. Why we're experiencing an AI boom: http://venturebeat.com/2016/08/12/why-is-now-the-time-for-artificial-intelligence/
21. Using ensembles of models to boost performance: http://www.datasciencecentral.com/profiles/blogs/improving-predictions-with-ensemble-model
22. AI in medical apps: https://www.datanami.com/2016/08/08/ai-initiative-targets-medical-apps/
23. Creating an invisible user interface with AI: https://techcrunch.com/2016/08/15/using-artificial-intelligence-to-create-invisible-ui/
24. Machine learning in finance-where it is, and where it's going: http://techemergence.com/machine-learning-in-finance-applications/
25. No matter how good the model, if it isn't interpreted or applied properly, it is useless: http://www.theverge.com/2016/8/19/12552384/chicago-heat-list-tool-failed-rand-test
26. Samsung demos its camera based on IBM's neuromorphic chip: http://www.extremetech.com/extreme/233747-samsung-demonstrates-camera-sensors-hooked-to-ibms-brain-imitating-silicon
27. An AI-based assistant for firefighters: http://www.jpl.nasa.gov/news/news.php?feature=6590
28. Deep learning is now being used in the Dragon speech recognition system: https://techcrunch.com/2016/08/16/dragon-15/
29. Companies continue to invest big in AI: http://www.theregister.co.uk/2016/08/17/chip_giants_invest_heavily_to_boost_changes_in_embedded_ai_platforms/?mt=1472434359193
30. OpenAI is set to receive the first "deep learning in a box" system: http://www.techrepublic.com/article/elon-musk-backed-openai-project-will-get-first-deep-learning-supercomputer-in-a-box/
31. Facebook open sources its fast text processing system: https://techcrunch.com/2016/08/18/facebooks-artificial-intelligence-research-lab-releases-open-source-fasttext-on-github/
32. How to train an AI doctor: http://www.datasciencecentral.com/profiles/blogs/training-an-ai-doctor-by-tyler-schnoebelen
33. Boosting your competitive advantage using machine learning: http://www.datasciencecentral.com/profiles/blogs/machine-learning-becomes-mainstream-how-to-increase-your
34. Mapping poverty using machine learning: http://motherboard.vice.com/en_au/read/artificial-intelligence-is-predicting-human-poverty-from-space
35. Machine learning systems now out-perform humans in diagnosing cancer biopsies: http://www.extremetech.com/extreme/233746-ai-beats-doctors-at-visual-diagnosis-observes-many-times-more-lung-cancer-signals
36. An Android malware detection system using machine learning: http://www.techrepublic.com/article/droidol-android-malware-detection-based-on-online-machine-learning/
37. A chatbot to help homeless people get government housing: http://uk.businessinsider.com/chatbot-helps-homeless-josh-browder-2016-8?utm_content=bufferb3a52&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer?r=US&IR=T
38. Tracking Beijing pickpockets with machine learning: http://www.economist.com/news/science-and-technology/21705296-artful-dodger-your-time-may-be-up-cutpurse-capers
39. Demystifying deep reinforcement learning: https://www.nervanasys.com/demystifying-deep-reinforcement-learning/
40. Getting up to speed on deep learning: https://medium.com/the-mission/up-to-speed-on-deep-learning-august-update-part-1-25afc11aea6b#.ikqxgmnp0
41. So, which professions fit these criteria the most? 6 signs your job is going to be automated: http://www.fastcompany.com/3062739/the-future-of-work/six-very-clear-signs-that-your-job-is-due-to-be-automated
42. "Academic clickbait" - the right choice of title for a paper can massively increase its reach: https://www.insidehighered.com/views/2016/08/24/review-article-using-clickbait-techniques-scholarly-titles
43. Baidu has open sourced their deep learning toolkit: http://www.theverge.com/2016/9/1/12725804/baidu-machine-learning-open-source-paddle
44. Predicting air quality in South Africa with machine learning: http://spectrum.ieee.org/energywise/energy/environment/tackling-air-quality-prediction-in-south-africa-with-machine-learning?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+IeeeSpectrum+%28IEEE+Spectrum%29&utm_content=FaceBook
45. Samsung is embedding neural networks in the chips for their new phones: http://www.theregister.co.uk/2016/08/22/samsung_m1_core/
46. Identifying regions of poverty from satelite images using machine learning: http://www.theverge.com/2016/8/18/12522764/poverty-measurement-satellite-algorithms-night-vs-day-imaging
47. Yandex is using machine learning to target users with less-annoying ads: https://techcrunch.com/2016/08/23/yandex-applies-ai-to-filter-annoying-ads-on-android-powered-by-user-reports/
48. On how important sleep is for resetting the brain: https://www.theguardian.com/science/2016/aug/23/sleep-resets-brain-connections-crucial-for-memory-and-learning-study-reveals
49. Research begins on using deep learning to segment cancerous tissue in scans: https://www.uclh.nhs.uk/News/Pages/Researchbeginstodeveloppioneeringtechnologytoplanradiotherapytreatment.aspx
50. Comparison of deep learning and AI: http://www.datasciencecentral.com/profiles/blogs/6448529:BlogPost:459267
51. Part 1 of a gentle introduction to TensorFlow: http://www.kdnuggets.com/2016/08/gentlest-introduction-tensorflow-part-1.html
52. Part 2 of a gentle introduction to TensorFlow: http://www.kdnuggets.com/2016/08/gentlest-introduction-tensorflow-part-2.html
53. Using GPU to turn a PC into a supercomputer: http://spectrum.ieee.org/tech-talk/computing/hardware/use-a-gpu-to-turn-a-pc-into-a-supercomputer?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+IeeeSpectrum+%28IEEE+Spectrum%29&utm_content=FaceBook
54. 10 programming career tips: http://www.techrepublic.com/article/10-programming-career-tips-for-students-and-new-professionals/
55. 5 requirements for a successful mobile app: http://www.techrepublic.com/article/five-things-your-branded-mobile-app-needs-to-succeed/
56. 10 ways to make your mobile app fail: http://www.informationweek.com/software/productivity-collaboration-apps/10-ways-to-doom-your-next-mobile-app/d/d-id/1326590?image_number=11
57. 9 ways bad managers drive good employees away: https://www.entrepreneur.com/article/249903
58. Microsoft is putting deep neural networks in a fridge: https://techcrunch.com/2016/09/02/microsoft-is-putting-cortana-machine-learning-in-a-fridge/
59. Some basic stuff - Why neural networks need activation functions: http://www.kdnuggets.com/2016/08/role-activation-function-neural-network.html
60. 10 Java machine learning libraries: http://www.datasciencecentral.com/profiles/blogs/25-java-machine-learning-tools-libraries
61. How is growing up around AI going to affect the next generation? https://techcrunch.com/2016/09/03/growing-up-in-generation-ai/
62. How convolutional neural networks work: http://www.kdnuggets.com/2016/08/brohrer-convolutional-neural-networks-explanation.html
63. How search at eBay got a boost from machine learning: https://www.datanami.com/2016/08/22/search-engines-get-machine-language-boost/
64. What one university did to graduate more women from computer science: http://www.techrepublic.com/article/what-universities-can-do-to-graduate-more-women-in-comp-sci-7-tips-from-harvey-mudd/
65. Sturgeon's law applies to AI as well - 90 % of them are crap: http://motherboard.vice.com/en_au/read/why-an-ai-judged-beauty-contest-picked-nearly-all-white-winners
66. Skin colour is not the only attribute distinguishing different populations - think the programmers forgot that: http://motherboard.vice.com/en_au/read/why-an-ai-judged-beauty-contest-picked-nearly-all-white-winners
67. PayPal is using machine learning to detect fraud. All based on open-source tools, too: http://www.americanbanker.com/news/bank-technology/how-paypal-is-taking-a-chance-on-ai-to-fight-fraud-1091068-1.html
68. Microsoft open sources its deep learning toolkit: http://blogs.microsoft.com/next/2016/01/25/microsoft-releases-cntk-its-open-source-deep-learning-toolkit-on-github/#sm.00002eomx1dc9fa2te714z6of4st4
69. Classifying cucumbers using deep neural networks: https://cloud.google.com/blog/big-data/2016/08/how-a-japanese-cucumber-farmer-is-using-deep-learning-and-tensorflow
70. A neural network based app to find local food: http://www.dealstreetasia.com/stories/russia-digest-neural-network-based-app-launched-52383/
71. Machine checking of statistics in published papers: http://motherboard.vice.com/en_au/read/scientists-are-worried-about-peer-review-by-algorithm-statcheck
72. 82 free data science e-books from O'Reilly: http://www.oreilly.com/data/free/archive.html
73. Classifying urban sounds using deep neural networks: http://aqibsaeed.github.io/2016-09-03-urban-sound-classification-part-1/
74. Why you should stop trying to multi-task: https://www.insidehighered.com/blogs/call-action-marketing-and-communications-higher-education/stop-multitasking
75. Text-to-speech using deep ANN to directly generate waveforms rather than sequences of syllables: https://deepmind.com/blog/wavenet-generative-model-raw-audio/
76. Some thoughts on how AI could manage your money for you: https://techcrunch.com/2016/09/08/ai-can-make-your-money-work-for-you/
77. Some military uses of machine learning: https://www.datanami.com/2016/09/12/pentagon-eyes-ai-battlefield/
78. Computer vision as a service: https://techcrunch.com/2016/09/12/restb-ai-offers-custom-computer-vision-as-a-service/
79. Overview of reinforcement learning: http://www.datasciencecentral.com/profiles/blogs/reinforcement-learning-and-ai
80. How Apple's wireless earbuds could lead to always-on AI: https://techcrunch.com/2016/09/13/apples-ai-if-by-air/
81. The dangers of relying on a supervised learning model when making decisions: http://www.kdnuggets.com/2016/09/deception-of-supervised-learning.html
82. The future of AI and machine learning: http://www.datasciencecentral.com/profiles/blogs/a-sneak-peek-at-the-future-of-artificial-intelligence-the-newest
83. Nvidia's low-power AI optimised computer: https://techcrunch.com/2016/09/13/nvidias-tiny-new-self-driving-ai-computer-sips-power/
84. Glad this kind of thing doesn't happen much in NZ. Since I'm in a mixed-race marriage myself, it's very troubling: https://www.insidehighered.com/news/2016/09/19/racist-chalk-messages-college-directed-against-presidents-children-and-diversity
85. Music mastering with machine learning: http://betakit.com/landr-wants-to-make-the-music-industry-more-accessible-with-machine-learning-platform/
86. AI and automation will eliminate 6% of jobs in the US by 2021: https://www.theguardian.com/technology/2016/sep/13/artificial-intelligence-robots-threat-jobs-forrester-report
87. An overview of decision trees: http://www.kdnuggets.com/2016/09/decision-trees-disastrous-overview.html
88. Description of the bagging approach used in random forests: http://www.kdnuggets.com/2016/09/reandom-forest-criminal-tutorial.html
89. Career opportunities in AI: http://www.datasciencecentral.com/profiles/blogs/finding-career-opportunities-in-ai
90. Learning machine learning in a year: http://www.kdnuggets.com/2016/09/machine-learning-year-total-noob-effective-practitioner.html
91. Microsoft wants to use AI to "solve" cancer: https://techcrunch.com/2016/09/20/microsoft-wants-to-crack-the-cancer-code-using-artificial-intelligence/
92. Number of jobs replaced by AI will be smaller than expected: http://www.techrepublic.com/article/why-the-number-of-jobs-that-will-be-replaced-by-robots-is-lower-than-you-think/
93. An AI that learned how to play deathmatch Doom from pixel data: https://techcrunch.com/2016/09/21/scientists-teach-machines-to-hunt-and-kill-humans-in-doom-deathmatch-mode/
94. Paper on the Doom deathmatch AI: https://arxiv.org/abs/1609.05521
95. Overview of 9 key papers in deep learning: http://www.kdnuggets.com/2016/09/9-key-deep-learning-papers-explained.html
96. Machine-generated peer reviews pass for the real thing: https://www.timeshighereducation.com/news/robot-written-reviews-fool-academics
97. How publish or perish selects for bad research: https://www.theguardian.com/science/2016/sep/21/cut-throat-academia-leads-to-natural-selection-of-bad-science-claims-study?CMP=share_btn_tw
98. FuzzyML is the first IEEE standard to come out of the Computational Intelligence Society: http://standardsinsight.com/ieee_company_detail/the-value-and-process-of-creating-standards-cis-develops-its-first-standard
99. Apparently it's possible to use TensorFlow on D-Wave's quantum computers: http://www.computerworld.com/article/3122512/computer-hardware/d-wave-plans-to-ship-a-2000-qubit-quantum-computer-in-17.html
100. Sounds a lot like an ecosystem of Darwinian bots: https://techcrunch.com/2016/09/16/bazillion-beings-are-ai-driven-bots-that-have-to-earn-their-keep-or-die/
101. Microsoft claims their ANN-based speech recognition system is the most accurate: http://www.theregister.co.uk/2016/09/15/microsoft_lowest_error_rate_ai_speech_recognition/
102. Why AI is booming now: http://www.nytimes.com/2016/09/19/technology/artificial-intelligence-software-is-booming-but-why-now.html?partner=IFTTT&_r=1
103. Looks like it isn't easy to make a living with open source AI: http://venturebeat.com/2016/09/24/machine-learning-startup-h2o-lays-off-10-of-employees/
104. Deep learning leads to more accurate processing of mammograms: http://futurism.com/artificial-intelligence-reads-mammograms-with-99-accuracy/
105. A neural network zoo: http://www.asimovinstitute.org/neural-network-zoo/
106. Tried to build a system to identify sepsis, ended up with a system that predicted deaths: https://www.buzzfeed.com/stephaniemlee/how-a-failed-hospital-algorithm-could-save-lives?utm_term=.sigva50LA#.njbNW6Ryp
107. Google open sources its image auto-captioning system, based on TensorFlow: http://www.zdnet.com/article/whats-in-that-photo-google-open-sources-caption-tool-in-tensorflow-that-can-tell-you/
108. I regularly tell my daughter to not do a PhD-instead, choose a career with better security than academia: https://www.insidehighered.com/advice/2016/09/23/faculty-member-no-longer-advises-her-students-go-academe
109. The limits of machine learning: http://nautil.us/blog/the-fundamental-limits-of-machine-learning
110. Building a robot with object recognition using TensorFlow: https://www.oreilly.com/learning/how-to-build-a-robot-that-sees-with-100-and-tensorflow
111. Of course our AI are going to be racist/sexist. Biased data leads to biased models. http://www.dailymail.co.uk/sciencetech/article-3808834/Are-making-AIs-racist-sexist-Researchers-warn-machines-learning-human-biases.html
112. Why is this still surprising to people? I was teaching my undergrad students this 16 years ago: http://boingboing.net/2016/09/06/weapons-of-math-destruction-i.html
113. k-Means vs Expectation-maximization clustering: http://www.kdnuggets.com/2016/09/comparing-clustering-techniques-concise-technical-overview.html
114. A pretty weak argument from the IEEE on why people shouldn't use Sci-Hub: http://theinstitute.ieee.org/blogs/blog/scihubs-free-articles-are-anything-but-free
115. Deep learning-based language translation: https://techcrunch.com/2016/09/27/google-unleashes-deep-learning-tech-on-language-with-neural-machine-translation/
116. Paper on deep learning-based translation: https://arxiv.org/pdf/1609.08144v1.pdf
117. Description of what machine learning actually is: http://techemergence.com/what-is-machine-learning/
118. Microsoft's move towards AI: https://techcrunch.com/2016/09/26/microsoft-ceo-satya-nadella-on-how-ai-will-transform-his-company/
119. What companies get wrong about machine learning: http://fortune.com/2016/09/27/machine-learning/
120. How Microsoft is integrating AI into Office 365: https://techcrunch.com/2016/09/26/microsoft-brings-new-ai-powered-features-to-office-365-and-dynamics-365/
121. The points made in this article are even more important now: http://www.bioone.org/doi/full/10.1641/0006-3568%282005%29055%5B0390%3APOP%5D2.0.CO%3B2
122. You'd be hard-pressed to find a research paper author opposed to Sci-Hub, because more downloads==more citations: http://theinstitute.ieee.org/blogs/blog/scihubs-free-articles-are-anything-but-free
123. Don't think authors-content creators-are opposed to Sci-Hub. Publishers are,it threatens their extortionate profits: http://theinstitute.ieee.org/blogs/blog/scihubs-free-articles-are-anything-but-free
124. We're doomed: https://techcrunch.com/2016/09/28/facebook-amazon-google-ibm-and-microsoft-come-together-to-create-historic-partnership-on-ai/
125. List of 15 tutorials on deep learning: http://www.datasciencecentral.com/profiles/blogs/15-deep-learning-tutorials
126. Is a Java deep learning library worth $3M? https://techcrunch.com/2016/09/28/skymind-raises-3m-to-bring-its-java-deep-learning-library-to-the-masses/ Would have thought an open source project could do it.
127. Google's cloud-based deep learning API is in beta: http://www.infoworld.com/article/3125095/artificial-intelligence/google-cloud-machine-learning-hits-public-beta-with-additions.html
128. More opportunities than risks when investing in AI: https://techcrunch.com/2016/09/24/investing-in-ai-offers-more-rewards-than-risks/
129. Investment advisers are betting big on machine learning: http://www.ifa.com.au/news/16855-advisers-look-to-ai-for-future-of-investment-survey-shows
130. Five safety problems with AI: https://openai.com/blog/concrete-ai-safety-problems/ Most of which seem concerned with the safety of the AI...
131. I don't think any human-created AI can be unbiased, all humans have some biases: http://motherboard.vice.com/en_au/read/to-make-ai-less-biased-give-it-a-worldview-racism-fairness-algorithm
132. Machine learning is seen as a savior for security: http://www.techrepublic.com/article/how-machine-learning-and-ai-will-save-the-entire-security-industry/
133. Reverse engineering cloud-based machine learning models: http://www.theregister.co.uk/2016/10/01/steal_this_brain/
134. Paper on reverse-engineering machine learning models: https://regmedia.co.uk/2016/09/30/sec16_paper_tramer.pdf
135. Predicting future human behaviour with deep learning: http://www.kdnuggets.com/2016/09/predicting-future-human-behavior-deep-learning.html
136. Yahoo has open sourced its porn-detecting convolutional neural network: https://techcrunch.com/2016/09/30/yahoo-open-sources-its-porn-detecting-neural-network/
137. Reading people's facial expressions using Google's Cloud Vision API: http://www.theregister.co.uk/2016/09/30/we_feel_your_pain_sometimes/
138. Why human curation still has a place among algorithmic organisation of information: https://www.theguardian.com/technology/2016/sep/30/age-of-algorithm-human-gatekeeper
139. Researchers claim that there is no inborn aptitude for programming, it can all be taught: http://www.theregister.co.uk/2016/09/28/geek_gene_denied/ Only studied 1 university
140. Looks like a primitive version of the personality constructs from Neuromancer: http://www.theverge.com/a/luka-artificial-intelligence-memorial-roman-mazurenko-bot
141. The next generation of neural networks will use spiking neurons: http://www.datasciencecentral.com/profiles/blogs/beyond-deep-learning-3rd-generation-neural-nets
142. We need a Data Mining Code of Ethics, to prevent this kind of sloppy work affecting the public: https://mic.com/articles/156286/crime-prediction-tool-pred-pol-only-amplifies-racially-biased-policing-study-shows#.c5TAsSbvF
143. Review of six cloud-based machine learning services: http://www.infoworld.com/article/3068519/artificial-intelligence/review-6-machine-learning-clouds.html#tk.ifw-infsb
144. What happens if you are not careful with building your model? You get a "weapon of math destruction" http://spectrum.ieee.org/tech-talk/computing/software/are-you-making-a-weapon-of-math-destruction
145. Yes, machine learning models can be sexist, classist, racist, otherwise biased if data used to train them is biased: https://techcrunch.com/2016/10/11/is-machine-learning-sexist/
146. What will people do when AI replaces all of the jobs? https://philipdodson.wordpress.com/2016/10/15/when-work-doesnt-exist-what-will-you-do/
147. AI-written poetry. It's really bad. But I expect it will get better: http://www.aipoem.com/easypoem/
148. Google investigates a framework to deal with bias in machine learning: https://thestack.com/world/2016/10/11/google-brain-machine-learning-discrimination/
149. Paper on equal opportunity in machine learning: https://drive.google.com/file/d/0B-wQVEjH9yuhanpyQjUwQS1JOTQ/view 
150. Paper on external memory for deep neural networks: http://www.nature.com/articles/nature20101.epdf?referrer_access_token=frOErGoDiMdDbAcGAiOinNRgN0jAjWel9jnR3ZoTv0MggmpDmwljGswxVdeocYSuWA28NTxakh-dRc-_0c4BVXvapExdTwoFcAqeeInLf9sHqUdOmQFGF_e6ZjH8WoY_s2ttYIgDzb9ecBAHMb7VcnxXLJau2ZJIZLecBqbtchd4IvmmzjDPLjuvFnm4y0x8eZ5IdVTyVGNaTcM3Oytc15GurUgTWnjmHuHAwVKQzv19W3Md8UgYuguYGAFdHPi54xevgLJXA24IrGZkk34C1--AjZNdv-Yw9QvDs3FG3_jhrH9nwBYCWpL82141wbWyFQ544nsEcPz6s1leHCs0zfn1R2kfAl-TsNWZAvLq7PjT_hGVe3U98Z-BzCkp3lKXOxbEGSbsZZ-RcD4MAgpme_2tR-RDEi6aBmRbt4QlR-U%3D&tracking_referrer=spectrum.ieee.org
151. How to compete in the age of AI: http://dataconomy.com/competing-age-of-ai/
152. Simple advice for getting an academic job: https://www.insidehighered.com/advice/2016/09/29/simple-advice-getting-job-academe-essay
153. Collective learning in a deep neural network: http://spectrum.ieee.org/automaton/robotics/artificial-intelligence/google-wants-robots-to-acquire-new-skills-by-learning-from-each-other?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+IeeeSpectrum+%28IEEE+Spectrum%29&utm_content=FaceBook
154. Are schools preparing students for the age of AI? https://www.theguardian.com/technology/2016/oct/12/schools-not-preparing-children-to-succeed-in-an-ai-future-mps-warn
155. Smart traffic lights make commuting more efficient: http://spectrum.ieee.org/cars-that-think/robotics/artificial-intelligence/pittsburgh-smart-traffic-signals-will-make-driving-less-boring?utm_source=SocialFlow&utm_medium=Facebook&utm_campaign=Social&utm_content=FaceBook
156. Myself, I'd prefer Sir Patrick Stewart: https://www.theguardian.com/technology/2016/oct/14/robert-downey-jr-mark-zuckerberg-digital-assistant-jarvis-iron-man
157. If a monkey can't own the copyright on a selfie it took, a computer shouldn't be able to own a patent: http://www.theregister.co.uk/2016/10/17/ai_computers_to_register_their_own_patents/
158. AI makes us better at what we are already good at: http://www.cio.com/article/3128870/analytics/artificial-intelligence-making-us-better-at-the-things-we-do-best.html
159. Overview of ANN and deep learning: http://www.kdnuggets.com/2016/10/artificial-intelligence-deep-learning-neural-networks-explained.html
160. Reducing patient no-shows with machine learning: https://www.datanami.com/2016/10/12/predictor-looks-reduce-patient-no-shows/
161. Promoting collaboration in software development using machine learning: https://www.datanami.com/2016/10/11/ai-platform-targets-coder-shortage/
162. Paper on using deep reinforcement learning to play first-person shooter games: https://arxiv.org/abs/1609.05521
163. Microsoft claims human-level speech recognition performance: http://www.theverge.com/2016/10/18/13326434/microsoft-speech-recognition-human-parity
164. The Royal Navy is investigating the use of AI in threat assessment: http://www.theregister.co.uk/2016/10/18/royal_navy_ai_software/
165. More on spiking ANN: http://www.datasciencecentral.com/profiles/blogs/more-on-3rd-generation-spiking-neural-nets
166. An AI corporate executive: http://www.theregister.co.uk/2016/1/19/ai_bot_will_guide_finnish_it_firm/
167. Using machine learning to detect expenses fraud: https://techcrunch.com/2016/10/18/internal-expense-fraud-is-next-on-machine-learnings-list/
168. Is Facebook slipping behind in AI? https://www.datanami.com/2016/10/18/facebook-ai-efforts-seen-lagging/
169. A role for machine learning in predictive analytics for marketing: http://techemergence.com/predictive-analytics-for-marketing/
170. 3 things to consider when deciding if a business needs AI: http://www.techrepublic.com/article/does-your-business-need-ai-consider-these-3-things/
171. Paper: The Mythos of Model Interpretability: https://arxiv.org/abs/1606.03490
172. How Airbnb makes use of machine learning: http://techemergence.com/airbnb-machine-learning-data-social-science-make-work/
173. How will AI deal with moral dilemmas? https://techcrunch.com/2016/10/19/ai-autonomous-cars-and-moral-dilemmas/
174. Intro to using neural networks in Python: http://www.kdnuggets.com/2016/10/beginners-guide-neural-networks-python-scikit-learn.html
175. How Uber uses machine learning in most of what it does: http://www.techrepublic.com/article/how-data-and-machine-learning-are-part-of-ubers-dna/
176. I remember seeing nonsensical conference papers well before we had autocorrect: https://www.theguardian.com/science/2016/oct/22/nonsense-paper-written-by-ios-autocomplete-accepted-for-conference
177. I suspect that this is a symptom of us approaching peak-hype for machine learning: http://www.theregister.co.uk/2016/10/21/machine_learning_craze_reaches_freelance_market/
178. Brief summary of Stephen Hawking's thoughts on AI: http://betanews.com/2016/10/21/artificial-intelligence-stephen-hawking/
179. Assessing Clinton & Trumps emotional intelligence with AI: https://www.fastcompany.com/3064863/election-2016/watch-this-ai-platform-assess-trumps-and-clintons-emotional-intelligence
180. Why analog computing is a good match with AI: http://spectrum.ieee.org/automaton/robotics/artificial-intelligence/analog-and-neuromorphic-chips-will-rule-robotic-age?utm_source=SocialFlow&utm_medium=Facebook&utm_campaign=Social&utm_content=FaceBook
181. Deep Fried Data: http://idlewords.com/talks/deep_fried_data.htm … How machine learning can make anything "taste" good
182. I always preferred the character of Chandler myself: https://www.theguardian.com/technology/2016/oct/20/joey-friends-virtual-digital-avatar-chatbot
183. The blind spot in AI research: http://www.nature.com/news/there-is-a-blind-spot-in-ai-research-1.20805
184. List of five free ebooks on machine learning: http://www.kdnuggets.com/2016/10/5-free-ebooks-machine-learning-career.html
185. Universities helping their staff with online profiles, but could it lead to unwanted uniformity? https://www.timeshighereducation.com/news/secret-shoppers-pimp-academics-online-profiles
186. A rather inspiring post by my old school-mate LenaRobinson: http://www.kiwigray.com/blog-kiwigray/2016/10/24/play-to-win-just-like-the-all-blacks
187. Why we need more diversity in AI: http://spectrum.ieee.org/tech-talk/at-work/tech-careers/computer-vision-leader-feifei-li-on-why-ai-needs-diversity?utm_source=SocialFlow&utm_medium=Facebook&utm_campaign=Social&utm_content=FaceBook
188. AI is not out to get us: https://www.scientificamerican.com/article/ai-is-not-out-to-get-us/
189. An AI "judge" that predicts the outcome of cases: https://www.theguardian.com/technology/2016/oct/24/artificial-intelligence-judge-university-college-london-computer-scientists - I wonder if they called it Dredd?
190. Awesome falsehoods - should be required reading for every programmer: https://github.com/kdeldycke/awesome-falsehood
191. Why advances in AI should be kept in the public eye: https://techcrunch.com/2016/10/23/advancements-in-artificial-intelligence-should-be-kept-in-the-public-eye/
192. TechCrunch explains WTF machine learning is: https://techcrunch.com/2016/10/23/wtf-is-machine-learning/ But article describes ANN rather than machine learning
193. No, I don't think that AI is going to end humanity: http://www.theregister.co.uk/2016/10/25/will_ai_spell_the_end_of_humanity_the_tech_industry_wants_you_to_think_so/
194. Overview of recurrent neural networks: http://www.datasciencecentral.com/profiles/blogs/recurrent-neural-nets-the-third-and-least-appreciated-leg-of-the-
195. Some sociological issues of intelligent machines: http://www.kdnuggets.com/2016/10/when-bow-down-machine-overlords.html
196. Overview of deep learning on GPU: http://www.datasciencecentral.com/profiles/blogs/accelerated-computing-and-deep-learning
197. IBM has made Watson AI available as a cloud service: http://www.techrepublic.com/article/ibm-says-new-watson-data-platform-will-bring-machine-learning-to-the-masses/
198. An AI for digesting research papers: https://techcrunch.com/2016/10/25/iris-ai-for-science/ Poor AI
199. Tutorial on implementing classification measures in Python: http://machinelearningmastery.com/implement-machine-learning-algorithm-performance-metrics-scratch-python/
200. Microsoft has released the next version of its Cognitive Toolkit to beta: https://techcrunch.com/2016/10/25/microsoft-launches-the-next-version-of-its-deep-learning-toolkit-into-beta/
201. Embedding AI in airport security scanners: https://www.theguardian.com/technology/2016/oct/25/airport-body-scanner-artificial-intelligence Badder than a bad thing that's very, very bad.
202. Even a flower-order business is betting on AI to increase it's business: http://computerworld.com/article/3135380/artificial-intelligence/1-800-flowers-wants-to-transform-its-business-with-ai.html
203. The impact of machine learning / AI on privacy: https://techcrunch.com/2016/10/26/the-darker-side-of-machine-learning/
204. Is AI going to create more jobs than it destroys? http://computerworld.com/article/3135081/artificial-intelligence/ai-and-robots-arent-gunning-for-your-job-white-house-economist-says.html
205. Pick your pastiche with deep learning: https://techcrunch.com/2016/10/26/deep-learning-tool-lets-you-pick-your-pastiche-mostly-monet-a-dab-of-dore-and-a-pinch-of-picasso/
206. AI-generated encryption: https://techcrunch.com/2016/10/28/googles-ai-creates-its-own-inhuman-encryption/
207. Paper on AI-generated encryption: https://arxiv.org/pdf/1610.06918v1.pdf
208. Identifying malicious URLs using machine learning: http://www.kdnuggets.com/2016/10/machine-learning-detect-malicious-urls.html
209. IBM says that in 5 years Watson will be behind every business & personal decision: http://computerworld.com/article/3135852/artificial-intelligence/ibm-in-5-years-watson-ai-will-be-behind-your-every-decision.html
210. Humans and machines together give better language understanding: http://dataconomy.com/ai-language-understanding/
211. Microsoft's Cognitive Toolkit is intended to bring machine learning to the masses: http://www.networkworld.com/article/3134877/application-development/microsoft-wants-to-bring-maching-learning-into-the-mainstream.html
212. What I look for when examining a postgraduate thesis: …http://computational-intelligence.blogspot.com/2016/10/examining-postgraduate-theses.html
213. An argument for why academics should not spend time on social media: https://www.timeshighereducation.com/blog/why-academics-should-not-make-time-social-media
214. A fairly comprehensive introduction to using neural networks in TensorFlow: https://www.analyticsvidhya.com/blog/2016/10/an-introduction-to-implementing-neural-networks-using-tensorflow
215. An argument that current AI is not good enough to justify a universal basic income: https://www.technologyreview.com/s/602747/todays-artificial-intelligence-does-not-justify-basic-income/?utm_campaign=internal&utm_medium=homepage&utm_source=top-stories_2
216. Generative adversarial neural networks: http://www.kdnuggets.com/2016/10/deep-learning-research-review-generative-adversarial-networks.html
217. Human intelligence + artificial intelligence is the future, not AI alone: https://techcrunch.com/2016/11/01/how-combined-human-and-computer-intelligence-will-redefine-jobs/
218. CCTV operators will be the next group made redundant by AI: http://www.theregister.co.uk/2016/11/02/nec_cctv_ai/
219. How to get good at R: http://www.kdnuggets.com/2016/11/data-science-101-good-at-r.html
220. A Bot-builder for non-programmers: https://techcrunch.com/2016/11/02/general-catalyst-backed-octane-will-make-you-a-bot/
221. 8 pitfalls for developers moving from R to Python: http://www.kdnuggets.com/2016/11/r-user-frustrating-learning-python.html
222. Automating customer complaints with machine learning: https://techcrunch.com/2016/11/02/resolver/
223. How Bayesian inference works: http://www.datasciencecentral.com/profiles/blogs/how-bayesian-inference-works
224. Microsoft Concept Graph: giving machines and AI common sense: https://techcrunch.com/2016/11/01/microsoft-strives-to-give-computers-common-sense-with-concept-graph/
225. Why AI and machine learning is hard: http://www.techrepublic.com/article/why-ai-and-machine-learning-are-so-hard-facebook-and-google-weigh-in/
226. A piece on my home town: http://www.nzherald.co.nz/travel/news/article.cfm?c_id=7&objectid=11739176
227. How machine learning is used in higher education: https://www.datanami.com/2016/11/01/data-analytics-higher-education/
228. How AI will transform business: http://blog.leadcrunch.com/ai-business-interviews-olin-hyde-ai-will-transform-businesses-as-profoundly-as-the-advent-of-the-internet?utm_campaign=Olin%20External%20Guest%20Blogs%20and%20Content&utm_content=41697510&utm_medium=social&utm_source=facebook
229. Even with AI entering the workforce, people are still needed: http://www.theverge.com/a/verge-2021/stacy-brown-philpot-taskrabbit-ceo-interview-ai-gig-economy
230. DeepMind is planning on taking on Starcraft 2 next: https://www.theguardian.com/technology/2016/nov/04/starcraft-ii-deepmind-game-ai
231. Paper on the good and the bad of using machine learning in higher education: https://www.newamerica.org/education-policy/policy-papers/promise-and-peril-predictive-analytics-higher-education/
232. There's more and more competition for fewer and fewer full-time permanent academic positions, so people are leaving: http://www.abc.net.au/news/2016-11-07/lack-of-funding-sees-scientists-leaving-labs-in-droves/7996604
233. An advance in automated lip reading accuracy, using deep learning over a fairly restricted data set: http://www.theverge.com/2016/11/7/13551210/ai-deep-learning-lip-reading-accuracy-oxford
234. Creating an unbiased model is hard as long as data sets are unbalanced: https://techcrunch.com/2016/11/07/why-its-so-hard-to-create-unbiased-artificial-intelligence/
235. When AI have "erotic" dreams (NSFW): https://open_nsfw.gitlab.io
236. Cleaning podcasts with deep learning: http://www.kdnuggets.com/2016/11/deep-learning-cleans-podcast-ahem-sounds.html
237. A brief and basic introduction to neural networks: http://www.kdnuggets.com/2016/11/quick-introduction-neural-networks.html
238. Using machine learning to count dugongs in drone images: https://techcrunch.com/2016/11/09/counting-endangered-sea-cows-is-hard-so-were-going-to-make-ai-do-it/
239. Classifying books by genre from their cover art, using deep learning: https://www.technologyreview.com/s/602807/deep-neural-network-learns-to-judge-books-by-their-covers/

Neural Networks, Volume 84, Pages 1-180, December 2016

1. Model-based reinforcement learning with dimension reduction
Author(s): Voot Tangkaratt, Jun Morimoto, Masashi Sugiyama
Pages: 1-16

2. From free energy to expected energy: Improving energy-based value function approximation in reinforcement learning
Author(s): Stefan Elfwing, Eiji Uchibe, Kenji Doya
Pages: 17-27

3. Piece-wise quadratic approximations of arbitrary error functions for fast and robust machine learning
Author(s): A.N. Gorban, E.M. Mirkes, A. Zinovyev
Pages: 28-38

4. Mean-square exponential input-to-state stability of delayed Cohen–Grossberg neural networks with Markovian switching based on vector Lyapunov functions
Author(s): Zhihong Li, Lei Liu, Quanxin Zhu
Pages: 39-46

5. image state estimation for memristive neural networks with time-varying delays: The discrete-time case
Author(s): Sanbo Ding, Zhanshan Wang, Jidong Wang, Huaguang Zhang
Pages: 47-56

6. Semi-supervised learning for ordinal Kernel Discriminant Analysis
Author(s): M. Pérez-Ortiz, P.A. Gutiérrez, M. Carbonero-Ruz, C. Hervás-Martínez
Pages: 57-66

7. Global synchronization of memristive neural networks subject to random disturbances via distributed pinning control
Author(s): Zhenyuan Guo, Shaofu Yang, Jun Wang
Pages: 67-79

8. Adaptive PID control based on orthogonal endocrine neural networks
Author(s): Miroslav B. Milovanović, Dragan S. Antić, Marko T. Milojković, Saša S. Nikolić, Staniša Lj. Perić, Miodrag D. Spasić
Pages: 80-90

9. Emergence of low noise frustrated states in E/I balanced neural networks
Author(s): I. Recio, J.J. Torres
Pages: 91-101

10. Delay-distribution-dependent image state estimation for delayed neural networks with image-dependent noises and fading channels
Author(s): Li Sheng, Zidong Wang, Engang Tian, Fuad E. Alsaadi
Pages: 102-112

11. A neurodynamic approach to convex optimization problems with general constraint
Author(s): Sitian Qin, Yadong Liu, Xiaoping Xue, Fuqiang Wang
Pages: 113-124

12. Multistability of complex-valued neural networks with discontinuous activation functions
Author(s): Jinling Liang, Weiqiang Gong, Tingwen Huang
Pages: 125-142

13. A Self-Organizing Incremental Neural Network based on local distribution learning
Author(s): Youlu Xing, Xiaofeng Shi, Furao Shen, Ke Zhou, Jinxi Zhao
Pages: 143-160

14. New results on exponential synchronization of memristor-based neural networks with discontinuous neuron activations
Author(s): Abdujelil Abdurahman, Haijun Jiang
Pages: 161-171

15. Coexistence and local image-stability of multiple equilibrium points for memristive neural networks with nonmonotonic piecewise linear activation functions and unbounded time-varying delays
Author(s): Xiaobing Nie, Wei Xing Zheng, Jinde Cao
Pages: 172-180

IEEE Transactions on Neural Networks and Learning Systems, Volume 27, Issue 11, November 2016

1. Decomposition Techniques for Multilayer Perceptron Training
Author: Luigi Grippo; Andrea Manno; Marco Sciandrone
Page(s): 2146 - 2159

2. Learning Robust and Discriminative Subspace With Low-Rank Constraints
Authors: Sheng Li; Yun Fu
Page(s): 2160 - 2173

3. Decentralized Dimensionality Reduction for Distributed Tensor Data Across Sensor Networks
Authors: Junli Liang; Guoyang Yu; Badong Chen; Minghua Zhao
Page(s): 2174 - 2186

4. Learning Transferred Weights From Co-Occurrence Data for Heterogeneous Transfer Learning
Authors: Liu Yang; Liping Jing; Jian Yu; Michael K. Ng
Page(s): 2187 - 2200

5. Multiple Representations-Based Face Sketch–Photo Synthesis
Authors: Chunlei Peng; Xinbo Gao; Nannan Wang; Dacheng Tao; Xuelong Li; Jie Li
Page(s): 2201 - 2215

6. RBoost: Label Noise-Robust Boosting Algorithm Based on a Nonconvex Loss Function and the Numerically Stable Base Learners
Authors: Qiguang Miao; Ying Cao; Ge Xia; Maoguo Gong; Jiachen Liu; Jianfeng Song
Page(s): 2216 - 2228

7. Estimating Sensorimotor Mapping From Stimuli to Behaviors to Infer C. elegans Movements by Neural Transmission Ability Through Connectome Databases
Authors: Cheng-Wei Li; Chung-Chuan Lo; Bor-Sen Chen
Page(s): 2229 - 2241

8. A Comparison of Algorithms for Learning Hidden Variables in Bayesian Factor Graphs in Reduced Normal Form
Authors: Francesco A. N. Palmieri
Page(s): 2242 - 2255

9. Sparse Bayesian Classification of EEG for Brain–Computer Interface
Authors: Yu Zhang; Guoxu Zhou; Jing Jin; Qibin Zhao; Xingyu Wang; Andrzej Cichocki
Page(s): 2256 - 2267

10. Robust Kernel Low-Rank Representation
Authors: Shijie Xiao; Mingkui Tan; Dong Xu; Zhao Yang Dong
Page(s): 2268 - 2281

11. Improving on Deterministic Approximate Bayesian Inferences for Mixture Distributions
Authors: Yohei Nakada
Page(s): 2282 - 2300

12. Optimizing Single-Trial EEG Classification by Stationary Matrix Logistic Regression in Brain–Computer Interface
Authors: Hong Zeng; Aiguo Song
Page(s): 2301 - 2313

13. Online Learning ARMA Controllers With Guaranteed Closed-Loop Stability
Authors: Savaş Şahin; Cüneyt Güzeliş
Page(s): 2314 - 2326

14. Feature Extraction Using Memristor Networks
Authors: Patrick M. Sheridan; Chao Du; Wei D. Lu
Page(s): 2327 - 2336

15. Exponential Stability and Stabilization of Delayed Memristive Neural Networks Based on Quadratic Convex Combination Method
Authors: Zhanshan Wang; Sanbo Ding; Zhanjun Huang; Huaguang Zhang
Page(s): 2337 - 2350

16. Detecting Wash Trade in Financial Market Using Digraphs and Dynamic Programming
Authors: Yi Cao; Yuhua Li; Sonya Coleman; Ammar Belatreche; Thomas Martin McGinnity
Page(s): 2351 - 2363

17. Multi-AUV Target Search Based on Bioinspired Neurodynamics Model in 3-D Underwater Environments
Authors: Xiang Cao; Daqi Zhu; Simon X. Yang
Page(s): 2364 - 2374

18. A Consistent Model for Lazzaro Winner-Take-All Circuit With Invariant Subthreshold Behavior
Authors: Ruxandra L. Costea; Corneliu A. Marinov
Page(s): 2375 - 2385

19. Asymptotically Stable Adaptive–Optimal Control Algorithm With Saturating Actuators and Relaxed Persistence of Excitation
Authors: Kyriakos G. Vamvoudakis; Marcio Fantini Miranda; João P. Hespanha
Page(s): 2386 - 2398

20. Identification of Nonlinear Spatiotemporal Dynamical Systems With Nonuniform Observations Using Reproducing-Kernel-Based Integral Least Square Regulation
Authors: Hanwen Ning; Guangyan Qing; Xingjian Jing
Page(s): 2399 - 2412

21. Echo State Networks With Orthogonal Pigeon-Inspired Optimization for Image Restoration
Authors: Haibin Duan; Xiaohua Wang
Page(s): 2413 - 2425

22. Group Component Analysis for Multiblock Data: Common and Individual Feature Extraction
Authors: Guoxu Zhou; Andrzej Cichocki; Yu Zhang; Danilo P. Mandic
Page(s): 2426 - 2439

23. Synchronization Control of Neural Networks With State-Dependent Coefficient Matrices
Authors: Junfeng Zhang; Xudong Zhao; Jun Huang
Page(s): 2440 - 2447

24. Efficient χ2 Kernel Linearization via Random Feature Maps
Authors: Xiao-Tong Yuan; Zhenzhen Wang; Jiankang Deng; Qingshan Liu
Page(s): 2448 - 2453

Examining Postgraduate Theses

I've examined a number of postgraduate theses by this point in my career. These are Doctoral and Master's theses from New Zealand and overseas institutions.While most of those theses have been a real pleasure to review, some have been real horrors. Even the ones I enjoyed examining often had errors in them. The errors that appear, though, tend to be the same kind of errors. That is, candidates for higher degrees are making the same errors even though they are from different institutions. So, I've written this post to discuss these errors, and how to avoid them. This post, then, describes what I look for in a thesis, and what I don't want to see in a thesis. Since I've mostly examined Doctoral and Master's theses, these are the focus of this post.

The Examination Process

This varies a bit according to the institution, but the general structure is the same. The process usually goes something like this:

• I am asked if I am interested in examining the thesis. I usually get sent the candidate's name, the title of the thesis and the thesis abstract.

• I say "Yes, I am interested" although I occasionally say "No" if the thesis is outside of my field of expertise.

• Some time later, I receive an examination pack. This usually contains things like:
• The thesis to be examined
• Institutional guidelines for examiners
• A marking sheet, where I make my recommendation and comments
• Other forms for payment of the honorarium and tax 

• I read the thesis several times, making comments on the pages each time.

• Using my comments, I write a report on the thesis and make a recommendation.

• There is sometimes an oral exam / viva held later, although more and more institutions seem to be moving away from those now.

Examiners usually get paid an honorarium, and as a New Zealand resident the honorarium I receive from a New Zealand institution has always had tax deducted from it. The size of the honorarium varies between institutions, but if you consider the time it takes to examine a thesis, it comes out at substantially less than minimum wage.

I prefer to receive the thesis as a PDF, as I find the examination process much easier when done electronically. I usually load the PDF onto my tablet, so I can get some examination work done during my daily commute on the train. When I submit my report and recommendation, I send the marked-up PDF with it.

Clarity

A thesis should be clear. Don't leave the hard mental work to the reader of the thesis! Lay everything out for them, especially why are you doing this? There is presumably a reason for doing the work you did, apart from "your supervisor told you to do it". The motivation for pursuing the research in the thesis should be laid out clearly and as early as possible.

The literature review should relevant to the topic of the thesis. I don't want to have to wade through pages of literature review that don't have anything to do with the thesis. Or, to put it another, way, don't "stuff" your literature review, it just annoys the examiner. A good question to ask about any part of the literature review is "why is this in the thesis?".

The literature review should be to-the-point. Anyone examining the thesis will have been carefully chosen and they will be experts in the field. Spending pages reviewing or describing material that a third-year student in the field should know is a waste of time and space. Better to just cite the key relevant papers and move on.

The literature review should also be critical. What are the holes in the literature? What is wrong with what has been previously published? What could have been done better? The work in a thesis should build upon what has gone before, it is incredibly rare that a thesis introduces an entirely new field.

If a thesis isn't clear, then it won't pass the examination. If you're lucky, then the examiner will give you enough detail to fix it and another examination. If not, then you will fail.

Citations

Any statement that is made should be backed up by data, or logical argument, or citation. In a thesis, most statements will be backed up by citation. This is especially true of the literature review.

Citations should be formatted correctly. Citations that are in-text are usually done something like (Smith, 1999). When the citation is referred to directly, it is something like "as in Smith (1999)". This is also the form used when the citation leads the sentence, for example "Smith (1999) said that...". While most authors nowadays will be using reference-management software, you should know how to use the software and not rely on the defaults.

This might seem like a small thing, but every time a reader comes across an incorrectly-formatted citation, it can break the flow of their reading. Break the flow of reading enough and the reader gets frustrated. That's not what you want when the reader is an examiner with the power to make the last three (or more) years of your life irrelevant.

Typos

Typos are a fact of life. Everyone makes mistakes while writing, but there are some things you can do to reduce the number of mistakes that make it through to the examiner.

Firstly, use a spell-checker. These are so straightforward to use now that there is no excuse for any incorrectly spelt words to appear in a thesis that is going for examination. However, relying on a spell-checker is also dangerous. Spell-checkers only tell you if a word is spelt incorrectly, they won't tell you if they are the wrong word to use. So, proof-reading is still essential.

Grammatical errors should also be checked for. While English has its quirks, these quirks must be known and dealt with. Small errors in grammar can completely change the meaning of a sentence. A common error is using incorrect tenses. For example, experiments reported in the thesis have been done, they are in the past, so use the past tense to refer to them.

Tables and Figures

Tables and figures are one of the most effective ways of presenting data, provided they are used appropriately and carefully. There are some common mistakes that you must avoid in tables and figures.

Firstly, do not use unnecessary precision in a table. If the table is presenting the area of city blocks, then presenting areas to the square millimetre is excessively precise.

Secondly, every column at least should be labelled. There are exceptions, of course, but it is important to consider whether the table could be understood without the labels. The rows and tables should also be in a logical structure, with related values grouped together.

The caption of a table or figure should be stand-alone, and should explain what the table or figure is showing. For tables, that means that column labels need to be described or defined. That is, the reader should be able to interpret the table or figure without having to refer to the main text of the work. This is because a table or figure often ends up being displayed on a different page to the explanation of the table, and having to flip back-and-forth between pages while trying to understand presented data is annoying. This can lead to some long captions.

For plots of data, be careful with legends and labels. These should be informative, not just some default like "x" on the x-axis. Again, the goal is clarity, as the purpose of a plot is to communicate to the reader.

There is, in my opinion, almost no situation under which a 3D plot makes sense. The 3D bar-charts in MS Excel are particularly bad and should not be used under any circumstances. 3D plots serve no purpose other than to show that the author knows how to make them. They do not make data clearer, but they are harder to accurately interpret.

Do not use line plots for discrete data. For example, a school has three terms per year, and students may commence their studies at the start of any of the terms. If we were to plot the number of students who commenced in each term across a period of several years, we would use a scatter plot, because the quantity being plotted (number of students) is discrete. A line plot would imply that the number of students who commenced in a particular term is different halfway through the term than it is at the start of the term. Since we've already established that students commence at the start of the term, this is plainly incorrect.

If presenting several different series of values on the same plot, then distinguish between them by making the point markers clearly different shapes. Do not rely on colour for this! There are two reasons why: Firstly, a non-trivial portion of the population are sufficiently colour-deficient that they will not be able to perceive the difference, especially between red and green; Secondly, a thesis will likely be printed in greyscale, which completely hides the colours.

Check the cross-referencing to tables and figures. I once examined a thesis where all of the cross-referencing was incorrect - the cross-references in the text referred to figures and tables that did not exist - which made the results all but impossible to interpret. If you use a package like LaTeX to write your thesis, and carefully check the error messages when compiling your document, this is not an issue. For other writing software, like MS Word, you need to be a bit more careful.

Finally, do not use the word "plot" in a caption for a plot, or "table" in the caption for a table. I know what a table is, and I know what a plot is. I don't need to be told.

Equations

Used properly, equations are an effective way of communicating complex concepts. It is very easy for equations to become opaque and uninformative. To avoid this, equations must be laid out carefully and consistently. Again LaTeX is good for this kind of thing, its equation tools are very powerful.

Every variable in an equation should be defined somewhere, ideally following the first equation in which it is used. Similarly, variables should not be re-used. A table of symbols can be helpful.

Experiments

You must understand your data. What process created it? What are the variables? What do the variables mean? What are the ranges of the variables? What are the scales of the variables? Are they nominal, ordinal, interval? Remember, just because something is expressed as a number, doesn't mean you can do arithmetic with it. Some statistics are invalid for some kinds of data, so a working knowledge of measurement theory and statistics is essential.

Some data sets will have hidden biases. These biases will influence any model that is built using the data and must therefore be accounted for. Remember, if you are using biased data to build a model, you will end up with a biased model.

The data must be represented in a logical way. Some models like neural networks can only handle discrete values like class labels if they are represented orthogonally. 

When evaluating the accuracy of a classification model, you must give some thought to the distribution of classes in the data set. If 90 % of the data in the data set are from one class, then it is quite simple to create a model that is 90 % accurate: it just classifies every example as the most common class. A simple percentage accuracy is not, therefore, very useful for evaluating the performance of your model.

A single partitioning of data is not going to give an accurate estimate of performance of any model. The standard approach, therefore, is to cross-validate over the data set, with a separate, independent, validation set held out (note that some sources call this the test set - the name given doesn't matter, as long as you use such an independent data set). If the data set is too small to use cross-validation, then jackknife over the data set instead. Or, you can bootstrap the data. The point is, there are several different approaches that can be used to produce statistically reliable results. These approaches are so simple, and well-known, that I consider not using them to be sufficient reason to reject a thesis: the candidate plainly does not have sufficient skill in the field to qualify for a higher degree.

The set-up of experiments must be described in detail, including the parameters of any algorithms used. The goal is for all experiments to be reproducible. The description should also include reasons for selecting any particular algorithm. There is always a reason, and if a candidate can't justify their choice of algorithm, then I do wonder whether they understand the state of the art enough to qualify as a professional researcher. There is always a reason for selecting an algorithm, even if it is really "Because my supervisor told me to use it".

If the thesis is presenting a new or improved algorithm, then is must be compared to existing algorithms. The choice of algorithms compared to should be justified. It is very easy to find an algorithm that performs so badly that it makes a new algorithm look good by comparison. Be clear about why an algorithm was selected.

All results should be subjected to an appropriate statistical analysis. Statistics show us what the numbers are trying to say. Statistics allows us to separate reality from our own prejudices. A good working knowledge of statistics is, therefore, extremely important.

The thesis should interpret the results for the reader. In other words, the thesis should explicitly answer the question "What do the results mean?". This interpretation, of course, must be done within the context of the statistical analysis. The results should also be compared to the literature where possible. Don't leave this interpretation up to the examiner! The examiner might not interpret things the way that you intended.

Response

When the examiners' reports are received by the institution, they will be collated and made available to the candidate. The candidate always has a right of response. Don't be afraid to disagree with an examiner! Examiners are human, they make mistakes, or they might have missed something in the literature. If a candidate does disagree, however, then they should have a solid justification for disagreeing. The candidate will have to convince the examiner that they were mistaken, that means using facts or logical argument. Personally, I am quite prepared to be proven wrong on anything I write in an examiner's report. But I will only be swayed by a convincing, well-reasoned argument based on either logic or data. If a candidate tries to bullsh*t me, then I will not react well.

If a viva is to be held, then the examiners' reports will be made available to the candidate well before. A viva is a way of demonstrating that the candidate really does know what they are talking about in the thesis, and that they are able to handle questions on their own. It is also an opportunity for the examiners to clarify any lingering issues from the examination. I never try to make a candidate feel uncomfortable or upset in the viva, and I don't understand examiners who do that. It is not an opportunity for an examiner to show off how clever they are, or to exercise their limited power over another person. It is the last step of the examination, and it should be carried out in a professional and collegial manner.

Summary

A postgraduate degree represents a substantial investment of time and effort on the part of the candidate and their supervisor. It behooves all involved in that process to minimise the chances that the effort will be wasted. Putting in the effort to avoid the common issues I have identified above will help to achieve this.

IEEE Transactions on Evolutionary Computation, Volume 20, Issue 5, October 2016

1. Pareto or Non-Pareto: Bi-Criterion Evolution in Multiobjective Optimization
Author(s): Miqing Li; Shengxiang Yang; Xiaohui Liu
Page(s): 645 - 665

2. An Analysis of the Inertia Weight Parameter for Binary Particle Swarm Optimization
Author(s): Jianhua Liu; Yi Mei; Xiaodong Li
Page(s): 666 - 681

3. Algebraic Differential Evolution Algorithm for the Permutation Flowshop Scheduling Problem With Total Flowtime Criterion
Author(s): Valentino Santucci; Marco Baioletti; Alfredo Milani
Page(s): 682 - 694

4. Kuhn–Munkres Parallel Genetic Algorithm for the Set Cover Problem and Its Application to Large-Scale Wireless Sensor Networks
Author(s): Xin-Yuan Zhang; Jun Zhang; Yue-Jiao Gong; Zhi-Hui Zhan; Wei-Neng Chen; Yun Li
Page(s): 695 - 710

5. A Hybrid Evolutionary Immune Algorithm for Multiobjective Optimization Problems
Author(s): Qiuzhen Lin; Jianyong Chen; Zhi-Hui Zhan; Wei-Neng Chen; Carlos A. Coello Coello; Yilong Yin; Chih-Min Lin; Jun Zhang
Page(s): 711 - 729

6. Metaheuristics for Specialization of a Segmentation Algorithm for Ultrasound Images
Author(s): Francesco Rogai; Claudia Manfredi; Leonardo Bocchi
Page(s): 730 - 741

7. On Routine Evolution of Complex Cellular Automata
Author(s): Michal Bidlo
Page(s): 742 - 754

8. Objective Extraction for Many-Objective Optimization Problems: Algorithm and Test Problems
Author(s): Yiu-ming Cheung; Fangqing Gu; Hai-Lin Liu
Page(s): 755 - 772

9. A Reference Vector Guided Evolutionary Algorithm for Many-Objective Optimization
Author(s): Ran Cheng; Yaochu Jin; Markus Olhofer; Bernhard Sendhoff
Page(s): 773 - 791

10. A Self-Organizing Multiobjective Evolutionary Algorithm
Author(s): Hu Zhang; Aimin Zhou; Shenmin Song; Qingfu Zhang; Xiao-Zhi Gao; Jun Zhang
Page(s): 792 - 806

11. Pareto Fronts of Many-Objective Degenerate Test Problems
Author(s): Hisao Ishibuchi; Hiroyuki Masuda; Yusuke Nojima
Page(s): 807 - 813

12. Stability Analysis of the Particle Swarm Optimization Without Stagnation Assumption
Author(s): Mohammad Reza Bonyadi; Zbigniew Michalewicz
Page(s): 814 - 819