Coronary heart disease diagnosis using kernel PCA and adaptive boosting

Amirhossein Koneshloo, Dongping Du

Research output: Contribution to conferencePaper

Abstract

Machine learning has been widely applied in combination with clinical practice guidelines for accurate cardiac diagnosis and prognosis. Abundant classification algorithms, e.g., Decision Tree, and Neural Network, have been developed to identify whether a patient has heart disease. It should be noted that the performance of the classification algorithms relies greatly on the features extracted from risk factors. The impact of nonlinear data transformation and dimensionality reduction on machine learning techniques have not been fully studied. In this investigation, we explore the nonlinear mapping of the risk factors into a higher dimension space, and then conduct the Principle Component Analysis (PCA) to find the largest possible variances, which are further used in the well-known adaptive learning algorithm i.e., ADABOOST, for classification. To evaluate the performance of the proposed method, a K-fold cross-validation is applied. The Cleveland Clinic Foundation heart disease dataset provided by UCI Machine Learning Repository is used for testing and validation. The accuracy obtained from the proposed strategy is 96.27%, which outperforms the models in the literature.

Original languageEnglish
Pages1157-1162
Number of pages6
StatePublished - 2018
Event2018 Institute of Industrial and Systems Engineers Annual Conference and Expo, IISE 2018 - Orlando, United States
Duration: May 19 2018May 22 2018

Conference

Conference2018 Institute of Industrial and Systems Engineers Annual Conference and Expo, IISE 2018
CountryUnited States
CityOrlando
Period05/19/1805/22/18

Keywords

  • Adaptive learning
  • Coronary heart disease
  • Diagnosis
  • Kernel PCA

Fingerprint Dive into the research topics of 'Coronary heart disease diagnosis using kernel PCA and adaptive boosting'. Together they form a unique fingerprint.

  • Cite this

    Koneshloo, A., & Du, D. (2018). Coronary heart disease diagnosis using kernel PCA and adaptive boosting. 1157-1162. Paper presented at 2018 Institute of Industrial and Systems Engineers Annual Conference and Expo, IISE 2018, Orlando, United States.