When numbers are bad, rejection is likely; despite reason.
ML related posts
Implementing sparse networks using neurons as building blocks.
Autograd is not a magic. It is a very simple idea implemented carefully
OpenMPI implementation of K-Means.
Training various Conv-Nets for the Trafic-Sign-Detection-Benchmark and comparing various hyper-parameter settings
A small review of literature on network pruning
Implementation of network pruning using torch.
Code to classify spam emails using lib-svm.