I like to write tutorials and notes about machine learning and learning systems.

  • Open Source Design Notes on Deep Learning Systems
    This is a series of notes on system design of deep learning systems. They can help general audiences understand the motivations, benefits and drawbacks of design choices in state-of-art deep learning. This will help deep learning practitioners as well as builders of other deep learning systems. In collaboration with DMLC/MXNet team members.

  • Introduction to Boosted Trees [Slides]
    This is a lecture on gradient boosting and boosted trees I made on Oct. 22, when I am TAing Machine learning in UW.
    I am taking the statistical view following the idea of LogitBoost, to present boosted trees algorithm as optimization for training loss and regularization in functional space. This is the model used in the XGBoost package

  • Introduction to Expression Template
    This is an tutorial about expression template. This is the trick that enables efficient and simple machine learning codes in C++