SHAP for explainable machine learning

I have always been very interested in explainability of algorithms, stemming from the curiosity of understanding how models work. I came to realize that the progress of machine learning is largely credited to the power of algorithms in capturing the delicate and complicated interactions between features. The most powerful of... [Read More]

Domain Adaptation - Feature Augmentation

After introducing the instance weighting framework for domain adaptation in the previous blog, I will explore a different framework. The second category of algorithms is feature-based. Instead of taking each instance as a whole, we try to extract useful information from the features. Here we explore three algorithms and applied... [Read More]

Domain Adaptation - Instance Weighting

Domain adaptation is an important area in transfer learning. The goal is grand: to deploy a model on a different domain from which it was trained on. A domain can be simply thought of as a different class of data. One of the examples is sentiment analysis on customer reviews... [Read More]

Machine learning - Week 2

This is the second full-time week spent on things besides my PhD and time has been distributed between algorithms and machine learning. Cramming the algorithms (from computer science) has surprised me about how little I know ‘behind the scene’. More importantly, I’ve become motivated to absorb powerful design principles and... [Read More]

Machine learning - Week 1

To get myself educated about this exciting field, I have started ‘cramming’ the fundamentals in the field. Although there are many cool and complicated tools in libraries and I can use them in a plug-and-go fashion, I do not find this approach satisfying. I want to know why things work.... [Read More]