Mustangs is an generative adversarial networks (GANs) training framework that combines E-GANs [1], which apply the principles of evolutionary computing to train GANs by generating diversity in terms of (gradient-based) mutations applied to the generator, and Lipizzaner [2], which uses a spatially distributed coevolutioary algorithms to optimize two populations of networks (generators and discriminators). Mustan mitigates problems such as instability and mode collapse during the training process.

Counterfactual regression (CFR) by learning balanced representations, as developed by Johansson, Shalit & Sontag (2016) and Shalit, Johansson & Sontag (2016). cfrnet is implemented in Python using TensorFlow 0.12.0-rc1 and NumPy 1.11.3. The code has not been tested with TensorFlow 1.0.

Accelergy is an energy estimation infrastructure for accelerator energy estimations.

Instruction THroughput Estimator using MAchine Learning (ITHEMAL)

Minimal, standalone library for solving GLMs in PyTorch

Efficient Labeled Data Factory with Minimal Human Effort

Measuring Generalization with Optimal Transport

NeuraCrypt encodes raw patient data using a randomly constructed neural network known only to the data-owner, and publishes both the encoded data and associated labels publicly. 

Brain Visualization Software.

ml4h is a toolkit for machine learning on clinical data of all kinds including genetics, labs, imaging, clinical notes, and more.