- https://en.wikipedia.org/wiki/Consumption_smoothing
- http://www.bobsguide.com/guide/news/2015/May/21/mifid-ii-controlling-and-testing-algorithms/
- http://www.seykota.com/Tribe/fractals/index.htm
https://www.udacity.com/course/intro-to-machine-learning--ud120
https://www.cs.ox.ac.uk/people/nando.defreitas/machinelearning/
Focuses on neural networks. This first two lectures are a great introduction to deep learning concepts and the second lecture in particular provides a list of background knowledge you should have.
Straightforward lectures explaining basic linear algebra concepts like matricies, vectors and scalar values.
https://www.youtube.com/watch?v=xyAuNHPsq-g&list=PLFD0EB975BA0CC1E0
These videos were helpful for understanding how to use data when training models, specifically how to think about the way your construction of input will influence your output.
- https://www.youtube.com/watch?v=Xh6Rex3ARjc
- https://www.youtube.com/watch?v=ksOErR1ldgo
- https://www.youtube.com/watch?v=P-WYkSZp9lY
- https://www.youtube.com/watch?v=eVSIDf5w1_I
Explains the underlying calculus through examples of Neural Networks in Python.
https://www.youtube.com/watch?v=bxe2T-V8XRs
Only parts 1 and 2 are out so far, but they are interesting and being able to use JS instead of Python let me get started much much faster.
https://hackernoon.com/machine-learning-with-javascript-part-1-9b97f3ed4fe5
There's an entire course based on using Natural for NLP. Clint found this, but I have not taken the course yet. https://egghead.io/courses/natural-language-processing-in-javascript-with-natural
- https://github.com/wooorm/retext-profanities
- https://github.com/wooorm/retext-sentiment
- https://github.com/wooorm/alex
- https://github.com/wooorm/retext-equality
When reseaching how to characterize suggestive language or innudendo rather than explicit language or profanity, I found this paper about identifying double entendre: http://people.cs.umass.edu/~brun/pubs/pubs/Kiddon11.pdf
It is implemented in this Ruby package with has an entertaining set of training data: https://github.com/sengupta/twss/blob/master/twss/positive.txt