Generative Adversarial Network Based Improved Transfer Learning Models for Chronic Heart Failure Detection
Bibliographic record
Abstract
Chronic heart failure (CHF) detection remains a critical challenge in healthcare due to its complex and multifactorial nature.Heart sound analysis serves an important part in identifying cardiovascular disease; however, the availability of balanced datasets for training machine learning models remains a challenge due to inherent class imbalances.To address this, an inception-based Generative Adversarial Networks (GAN) method is proposed to acquire the distribution of heart sound classes and generate synthetic samples for underrepresented classes.The model is applied to the unbalanced PhysioNet dataset of heart sound signals, and features extracted from real and synthetic data are combined into feature vectors, enabling feature fusion, which are passed to various classification models.This study attempts to fine-tune pre-trained convolutional neural network models, specifically VGG16 and MobileNet for classification of Heart sound signals.The results of proposed models are compared with and without GAN model on heart sound signals and gets significant improvement with KNN Hyper parameter tuning, Proposed Autoencoder + CNN model and Fine-tune MobileNet and VGG16 algorithm.KNN hyperparameter tuning refines the model's decision boundaries for better classification accuracy, while the Autoencoder + CNN architecture leverages deep feature learning to extract high-level representations, enhancing diagnostic precision.The model outperforms machine learning and deep learning models, improving overall recall and F1-score by approximately 8%.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".