Python Machine Learning Framework which aims fast access to many standard datasets (like MNIST, CIFAR, NIPS, yahoo_ltrc1, medical and etc).
This repository is a migration from BitBucket of Hugo Larochelle: MLPython in order to maintain more developer friendly environment, better documentation and pip packaging.
For quick installation with pip:
pip install mlpython
ML python also uses some native codes in C programming language for correct installation please refer to : Installation
import mlpython.datasets.store as dataset_store
import os
os.environ["MLPYTHON_DATASET_REPO"] = "/home/mylinux_home/directory"
dataset_store.download('mnist')
trainset,validset,testset = dataset_store.get_classification_problem('mnist')
For the full documentation pleas check here: