Multi Layer Perceptron written in Python3 with scikit-lean
pip install git+https://github.com/masaponto/python-mlp
- scikit-learn: 0.17.1 and up
- numpy: 1.10.4 and up
from mlp import MLP
from sklearn.datasets import fetch_mldata
from sklearn.preprocessing import normalize
from sklearn.model_selection import train_test_split
db_name = 'iris'
data_set = fetch_mldata(db_name)
data_set.data = normalize(data_set.data)
mlp = MLP(hid_nums=[5], epochs=1000, batch_size=1)
X_train, X_test, y_train, y_test = train_test_split(
data_set.data, data_set.target, test_size=0.4)
mlp.fit(X_train, y_train)
print("Accuracy %0.3f " % mlp.score(X_test, y_test))MIT