Breast cancer diagnosis using a hybrid genetic algorithm for feature selection based on mutual information

Alzubaidi, A ORCID: https://orcid.org/0000-0002-5977-564X, Cosma, G ORCID: https://orcid.org/0000-0002-4663-6907, Brown, D ORCID: https://orcid.org/0000-0002-1677-7485 and Pockley, AG ORCID: https://orcid.org/0000-0001-9593-6431, 2016. Breast cancer diagnosis using a hybrid genetic algorithm for feature selection based on mutual information. In: Proceedings: iTAG 2016: the 2016 International Conference on Interactive Technologies and Games - EduRob in Conjuction with ITAG2016 - 26-27 October 2016, Nottingham, United Kingdom. Washington, DC: IEEE Computer Society Conference Publishing Services, pp. 70-76. ISBN 9781509037384

Preview

Text PubSub6981_Cosma.pdf - Post-print Download (651kB) | Preview

Abstract

Feature Selection is the process of selecting a subset
of relevant features (i.e. predictors) for use in the construction of predictive models. This paper proposes a hybrid feature selection approach to breast cancer diagnosis which combines a Genetic Algorithm (GA) with Mutual Information (MI) for selecting the best combination of cancer predictors, with maximal discriminative capability. The selected features are then input into a classifier to predict whether a patient has breast cancer. Using a publicly available breast cancer dataset, experiments were performed to evaluate the performance of the Genetic Algorithm based on the Mutual Information approach with two different machine learning classifiers, namely the k-Nearest Neighbor (KNN), and Support vector machine (SVM), each tuned using different distance measures and kernel functions, respectively.
The results revealed that the proposed hybrid approach is highly accurate for predicting breast cancer, and it is very promising for predicting other cancers using clinical data.

Item Type:	Chapter in book
Creators:	Alzubaidi, A., Cosma, G., Brown, D. and Pockley, A.G.
Publisher:	IEEE Computer Society Conference Publishing Services
Place of Publication:	Washington, DC
Date:	15 December 2016
ISBN:	9781509037384
Identifiers:	Number Type 10.1109/iTAG.2016.18 DOI
Rights:	© 2016 IEEE.
Divisions:	Schools > School of Science and Technology
Record created by:	Jill Tomkinson
Date Added:	25 Jan 2017 10:39
Last Modified:	27 Aug 2021 09:56
URI:	https://irep.ntu.ac.uk/id/eprint/30008

Actions (login required)

Edit View

Statistics

Views

Downloads