/mnist

Some different approaches to the MNIST problem

Primary LanguageJupyter Notebook

A place to play with various methods for analyzing the MNIST data set

I'm not going to host the DBs here - check out http://yann.lecun.com/exdb/mnist/ for the data sets

Goals: to work on MNIST with many of the classifiers in sklearn, to learn which ones work well for this problem and which do not. I'd also like to learn about the parameters and options associated with the classifiers and learn about things like model selection and hyperparamter optimization.