About Programming

As a statistician, I value computational skills as well as theoretical skills. Coding algorithms helps me to understand the algorithm better than just theoretically thinking and reading from the paper. Here I share some of the useful tools I use to facilitate my study. 

"Journal papers are often limited to 8 pages and the authors often have to skip nitty-gritty algorithm details, and code optimizations. Reading about an algorithm in theory and coding it in practice are two different ball games. A good practice while coding computer vision algorithms is to have phantom images and ground truth, so you know for a fact that your algorithm is coded correctly. These phantom images can be artificially generated such as drawing a rectangle in Matlab and saving it. Then run it on your real dataset to see the performance of the algorithm." -- Ankur Kumar in Quora


Code Practice

LeetCode [source]

Topcoder [source]

Careercup [source]

MODE-SQL [source]

HackerRank [source]

GeeksforGeeks [source]

Deep Learning

TensorFlow [source]

End2end DL [source]

Chrome Music Lab [source]

Magenta [source]

Financial Market

Option, Future and Other Derivatives [book]

Other Sources

Paper with Code [source]

NLP Paper with Code [source] 

Big-O Cheat Sheet [source]

Acing AI Interviews [source]

Private Resources [source]



My Career Choice

Dream Job!

  • Get a doctorate in top academic institute
  • Get postdoc training with top scholars
  • Publish high quality paper
  • Apply for funding from NSF
  • Academic social networking
  • Read BOOKs!

Professor

To be continued.

2nd Choice

  • Research Scientist in universities
  • Research Scientist in high-tech companies

Research Scientist

To be continued.

3rd Choice

Prefer high-tech companies in Silicon Valley.

  • FAIR
  • Google Brain 
  • DeepMind
  • Oracle
  • Amazon
  • Microsoft
  • LinkedIn
  • etc.

Data Scientist

To be continued.

3rd Choice

Prefer Wall Street hedge fund companies.


Quantitative Researcher (Quant)

I don't like the financial market and its risk at all. The only reason I would go for this position is the application part of mathematical and statistical methods. Somehow quant position gives me an impression of demanding more intellectual capacity and being more technical than Data Scientist. As a technical and mathematical person, I like to investigate data, and find evidence towards the answer. By the way, I'm a big fan of Detective Conan

Please contact me if you want to share some experience or have different insights.


Chicken Soup

The career advice I wish I had at 25, Shane Rodgers  [link]

Using Format