Feature Selection With Optimal Coordinate Ascent (OCA)

14 Pages Posted: 20 Dec 2018

See all articles by David Saltiel

David Saltiel

A.I. Square Connect

Eric Benhamou

A.I. Square Connect; LAMSADE- Paris Dauphine University

Date Written: December 3, 2018


In machine learning, Feature Selection (FS) is a major part of efficient algorithm. It fuels the algorithm and is the starting block for our prediction. In this paper, we present a new method, called Optimal Coordinate Ascent (OCA) that allows us selecting features among block and individual features. OCA relies on coordinate ascent to find an optimal solution for gradient boosting methods score (number of correctly classified samples). OCA takes into account the notion of dependencies between variables forming blocks in our optimization. The coordinate ascent optimization solves the issue of the NP hard original problem where the number of combinations rapidly explode making a grid search unfeasible. It reduces considerably the number of iterations changing this NP hard problem into a polynomial search one. OCA brings substantial differences and improvements compared to previous coordinate ascent feature selection method: we group variables into block and individual variables instead of a binary selection. Our initial guess is based on the k-best group variables making our initial point more robust. We also introduced new stopping criteria making our optimization faster. We compare these two methods on our data set. We found that our method outperforms the initial one. We also compare our method to the Recursive Feature Elimination (RFE) method and find that OCA leads to the minimum feature set with the highest score. This is a nice byproduct of our method as it provides empirically the most compact data set with optimal performance.

Keywords: feature selection, coordinate ascent, gradient boosting method

JEL Classification: C

Suggested Citation

Saltiel, David and Benhamou, Eric, Feature Selection With Optimal Coordinate Ascent (OCA) (December 3, 2018). Available at SSRN: https://ssrn.com/abstract=3293503 or http://dx.doi.org/10.2139/ssrn.3293503

David Saltiel (Contact Author)

A.I. Square Connect ( email )

35 Boulevard d'Inkermann
Neuilly sur Seine, 92200

Eric Benhamou

A.I. Square Connect ( email )

35 Boulevard d'Inkermann
Neuilly sur Seine, 92200

LAMSADE- Paris Dauphine University ( email )

Place du Marechal de Lattre de Tassigny
Pais, 75016

HOME PAGE: http://https://www.lamsade.dauphine.fr/

Register to save articles to
your library


Paper statistics

Abstract Views
PlumX Metrics