Saturday, February 13, 2016

Weekly Review 12 February 2016

Some interesting links that I Tweeted about in the last week:

  1. Financial robo-advisors: http://www.bloomberg.com/news/articles/2016-02-05/the-rich-are-already-using-robo-advisers-and-that-scares-banks
  2. A low-power neural network chip for deep learning http://news.mit.edu/2016/neural-chip-artificial-intelligence-mobile-devices-0203 
  3. Emphasis on metrics is harmful to universities http://www.theguardian.com/higher-education-network/2015/nov/27/our-obsession-with-metrics-turns-academics-into-data-drones
  4. Four take-aways about Google's Tensor Flow http://www.infoworld.com/article/3003920/data-science/4-no-bull-takeaways-about-googles-machine-learning-project.html
  5. The damage done by bad PhD supervisors http://www.theguardian.com/higher-education-network/2015/dec/11/bad-phd-supervisors-can-ruin-research-so-why-arent-they-accountable Bad post-doc supervisors can ruin careers, too. I've seen it happen.
  6. AI for network security analysis http://www.net-security.org/secworld.php?id=19409
  7. Some researchers have used my SECoS algorithm to do this: http://www.net-security.org/secworld.php?id=19409 
  8. Version 1.22 of the ECoS Toolbox has been released: http://ecos.watts.net.nz/Software/Toolbox.html 
  9. SECoS compiler tool now outputs Java and PHP code, in addition to C++, C# and Python http://ecos.watts.net.nz/Software/Toolbox.html
  10. Added tools for analysing trained NECoS networks to the ECoS Toolbox version 1.22 http://ecos.watts.net.nz/Software/Toolbox.html
  11. Implementing k-nearest neighbor algorithm using Python: http://blog.cambridgecoding.com/2016/01/16/machine-learning-under-the-hood-writing-your-own-k-nearest-neighbour-algorithm/
  12. Using email effectively https://www.insidehighered.com/advice/2016/02/05/how-use-email-more-effectively-essay I try to have an empty (work) inbox when I go home in the evening
  13. Grouping coyote, wolf howls into "dialects" using machine learning http://motherboard.vice.com/read/wolves-have-different-howling-dialects-machine-learning-finds
  14. Canine howl dialects, paper here http://www.sciencedirect.com/science/article/pii/S0376635716300067
  15. Common data visualisation mistakes: http://www.kdnuggets.com/2016/02/common-data-visualization-mistakes.html
  16. List of 34 machine learning resources and articles: http://www.datasciencecentral.com/profiles/blogs/34-external-machine-learning-resources-and-related-articles
  17. Machine learning in dating websites: http://www.kdnuggets.com/2016/02/does-machine-learning-allow-opposites-attract.html
  18.  Image editing with Python: http://motherboard.vice.com/en_au/read/hack-this-edit-an-image-with-python 
  19. Artificial intelligence and sarcasm: https://thestack.com/cloud/2016/02/11/why-sarcasm-is-such-a-problem-in-artificial-intelligence/
  20. Predicting cancer survival with machine learning: http://www.infoq.com/articles/health-informatics-apache-spark-machine-learning 
  21. A Gentle Guide to Machine Learning: https://blog.monkeylearn.com/a-gentle-guide-to-machine-learning/ 
  22. Time-Series prediction with Python: http://www.analyticsvidhya.com/blog/2016/02/time-series-forecasting-codes-python/ 
  23. Women are still badly under-represented in tech https://medium.com/@lkr/i-m-a-woman-in-tech-but-even-i-didn-t-get-it-until-this-week-350cf8b62c46#.8bdgh633h I've increased the number of women teaching in my department.
  24. Four myths about using social media as a researcher: http://www.fasttrackimpact.com/#!The-four-greatest-myths-about-using-social-media-as-a-researcher/hmlp3/56a22b2f0cf2009838b71c56