quicknn 1.0.1

A package for quick&dirty DNN

QuickDNN

An implementation sklearn-like of FeedForward Neural Networks for quick applications. It can handle categorical variables with one-hot-encoding(OHE) method batch-wise as well as continuous variables.

Example

from quickdnn import QuickDNN

# load X_train, X_test, y_train, y_test

qdnn = QuickDNN(list_neurons=[100, 100, 1]) # two hidden layers with 100 neurons each and 1 output neuron. 
qdnn.fit(X_train, y_train) # training phase
y_pred = qdnn.predict(X_test) # predicting phase

Installing

$ pip install quickdnn

License

This project is licensed under the MIT License - see the LICENSE.md file for details.