Name: Mike Hughes
Type: User
Company: @tufts-ml : Tufts University, Department of Computer Science
Bio: Assistant Prof. of Computer Science at Tufts @tufts-ml -- Postdoc Harvard SEAS @dtak (2016-18) -- Ph.D. Brown U. CS (2010-16) @bnpy -- B.S. Olin College (2010)
Location: Somerville, MA
Blog: https://www.michaelchughes.com
Mike Hughes's Projects
Experimental class-conditional Gaussian models.
Side-by-side comparison of machine learning algorithms in matlab and python.
Public-facing repo for code related to in-class demos and homework for CS 136 at Tufts in Spring '24
Implements common transforms that map constrained parameter spaces (positive reals, unit interval, unit simplex, cholesky space, etc) to unconstrained real values. Useful for gradient-based optimization.
The course website for EME185: http://moorepants.github.io/eme185/
Easy TOC creation for GitHub README.md
Archive of old course project on analyzing acoustic signals from computer keyboards. Unmaintained.
Fast, vectorized C++ implementation of K-Means using the Eigen matrix template library. Includes Matlab and Python interfaces.
Small experiments with comparing different logistic stick-breaking models for count data.
Source materials for my personal website
Specs for "mini projects" for candidate students interested in my research group
Draft website, using Pelican
Annotated dataset of simple human exercises (jumping jacks/arm circles/etc) recorded as motion capture traces.
Nonparametric Bayesian Inference for Sequential Data. Includes state-of-the-art MCMC inference for Beta process Hidden Markov Models (BP-HMM). Implemented in Matlab.
Python code to compare reidentification of subjects via demographics alone using NHANES 2003-04 dataset.
Example code for processing rate-limited remote API queries fast in parallel with asyncio
Prototype python code for object recognition using satellite images. Early-stage active development.
Supervised Latent Dirichlet Allocation and other topic models. Supports regression and classification. Written in Matlab.