Modelrithm

Python Framework that compares 7 different Machine Learning algorithms' accuracy, precision, and F-Beta scores efficiently (~0.15 second runtime) for a given user's training and testing data, and returns the highest scoring of each, along with automated hyperparameter optimization.

Installation

pip install Modelrithm

Requirements

• sklearn
• matplotlib
• numpy

Usage

from Modelrithm import Modelrithm

model = Modelrithm(X_test, Y_train, X_test, Y_train)
model.Classification()

Examples

• Aided in increasing accuracy, precision, and f-beta score of a classification problem using satellite pictures of the earth and moon.

  • This application is included under the 'Examples' folder

Accolades

Winner, Lockheed Martin Aerospace Challenge - SBHacks 2017

TODO

• Hyperparameter optimization by Random Search (not Grid Search, takes too long).