| Autor: |
Singh, Rajeswari Chhual, Maharana, Santosh Kumar, Mohapatra, Subasish |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2023, Vol. 2745 Issue 1, p1-9, 9p |
| Abstrakt: |
The availability of relevant and validated data is vital for many implementing machine learning approaches. Larger data size has negative impacts on the performance of the learning process. The proposed work desires to develop a strategy that stores a lot of unique facts communed to data with few features. Feature selection is a crucial predictor of classification to remove unnecessary features from a particular dataset. In general, feature selection can be assumed an issue to develop multiple objectives, i.e., to lessen the number of features and improve the classifiers' accuracy. The main aim of the feature selection process is to select the features which are affecting the target variable much. Considering the feature selection concern in healthcare, this work contributes a hybrid GA-based feature selection method to increase the classifier accuracy. All the genetic operator like crossover and mutation is applied to converge the solution towards the optimal result. Finally, the performance is measured using different parameters like accuracy, precision, recall, and F_measure. The proposed method result will be compared with existing machine learning classifiers to show the model's effectiveness. The result shows the proposed model achieves 97.66% accuracy and 98.28% F1_Score. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
