Instance Based Classifier on time series and static data

Interactive Weak Supervision: Learning Useful Heuristics for Data Labeling

kmeans clustering with a maximum cluster size constraint

Data analysis on Social Leader killings in Colombia in order to inform an investigative journalism piece.

Compute Adjusted Rand Index and Confidence Interval. The confidence interval is computed using central limit theorem and the exact variance of the Adjusted Rand Index.

Machine learning methods for clustering time series data in public policy.

A simple implementation of adagrad

Compute fast connected components for text data using MinHash

UI for the Image Query Adaptation (IQA) tool

Rest api for the Image Query Adaptation (IQA) tool

Active Learning using open source scikit tools

Matlab code for regularized metric learning using trace norm and frobenius norm regularization.

basic Bayesian optimization for text classification

Transfering Knowledge in Low-Rank Similarity Metric Learning

Scalable Domain Adaptive Similarity functions