Current coursework

I am currently taking a course on machine learning and trading. The syllabus can be found here. The course serves as an introduction to machine learning methods and how to apply them to finance.

Topics to discuss in course:

  • stock market basics
  • pandas
  • k nearest neighbor learning
  • decision trees
  • reinforcement learning

CS231n Mod1 part 1

See link here

What are some takeaways after reading this?

  • Computers see images as matrices of numbers
  • An image is described as a 2 dimensional matrix of pixel values
  • Each pixel represents color values
  • Distance between images is sum of difference of pixels

$$d_1 (I_1, I_2) = \sum_{p} \left| I^p_1 - I^p_2 \right| $$

Maintaining a bibliography

Something I should have done when I started grad school is maintain a bibliography. Throughout college and high school I used EasyBib or some other simple software to generate citations. Google Scholar will also generate citations in multiple formats, but accuracy varies.

There are software available to make data entry easier. I tried JabRef, which is a JVM based citation manager. JabRef compiles bibtex files. The interface is reasonably designed. Then I tried Zotero, which features web browser integrations. I like Zotero, the gtk gui looks nice and runs smoothly. The web browser plugin makes submitting entries with metadata fields filled in much easier. Zotero is the most convenient choice.

I will be storing my bibliographies in my practice git repository here