Feature Selection using a Genetic Algorithm for the Detection of Abnormal ECG Recordings

摘要

A new classification algorithm, called CFI (for Classification on, Feature Intervals), is developed and applied to problem of detecting abnormal ECG signals. The domain contains records of patients with known diagnosis. Given a training set of such records the CFI learns how to detect arrhythmia. CFI represents a concept in the form of feature intervals on each feature dimension separately. Classification in the CFI algorithm is based on a real-valued voting. A genetic algorithm is used to select the set of relevant features. Each selected feature equally participates in the voting process and the class that receives the maximum amount of votes is declared to be the predicted class. The performance of the CFI classifier is evaluated empirically in terms of classification accuracy and running time.