Evolutionary Feature Synthesis

Evolutionary feature Synthesis is a machine learning technique that evolves the features of a linear model such that the score of a model is maximized. For regression this is the r^2 values and for classification this is accuracy. Evolutionary Feature Synthesis has been shown to be competitive with the state of the art in machine learning, while retaining all the nice properties of linear models such as a convex error function and interpretability. For details refer to the paper.

Installation

efs is compatible with Python 2.7+

pip install efs 

Example Usage

import matplotlib.pyplot as plt
import numpy as np
from efs.evolutionary_feature_synthesis import EFSRegressor
from sklearn.model_selection import train_test_split

def target(x):
    return x**3 + x**2 + x 

Now we'll generate some data on the domain [-10, 10].

X = np.linspace(-10, 10, 100, endpoint=True)
y = target(X)
X_train, X_test, y_train, y_test = train_test_split(X, y)

Finally we'll create and fit the EFSRegressor estimator and check the score.

sr = EFSRegressor()
sr.fit(X_train, y_train)
score = sr.score(X_test, y_test)
print('Score: ' + score)