MobileNetV3: an efficient deep learning-based feature selection and classification technique for cardiovascular disease
Abstract Accurately identifying cardiovascular illness is one of the most important and difficult responsibilities in treating a patient before a heart attack. Unfortunately, most currently utilized cardiovascular disease prediction algorithms could not achieve higher accuracy due to inadequate fore...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-07-01
|
| Series: | Journal of Engineering and Applied Science |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s44147-025-00654-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Accurately identifying cardiovascular illness is one of the most important and difficult responsibilities in treating a patient before a heart attack. Unfortunately, most currently utilized cardiovascular disease prediction algorithms could not achieve higher accuracy due to inadequate forecasting methodology and data-recognized techniques. This research proposes a novel Deep Learning (DL) based MobileNetV3 method for categorizing cardiovascular disease, driven by the current obstacles. Missing data handling, outlier detection, normalization using min–max normalization methods, categorical data encoding, and transformation are all done during the pre-processing phase. Next, use the EfficientNetV2 approach to extract the features from the data, and then use the Reptile Search Algorithm (RSA) to choose the most significant features from the collected features. Lastly, use the MobileNetV3 approach to categorize cardiovascular illness based on specific criteria. The proposed MobileNetV3 approach uses benchmark datasets to predict the patients’ cardiovascular illness with 99.82% and 99.73% accuracy for the UCI Heart Disease and Framingham datasets, respectively. The simulation findings indicate that the proposed MobileNetV3 strategy outperforms standard learning techniques based on multiple performance matrices. |
|---|---|
| ISSN: | 1110-1903 2536-9512 |