🫀 ECG Classification with Neural Networks
Accurate classification of electrocardiogram (ECG) signals is essential for diagnosing cardiovascular diseases and supporting timely medical interventions. This project explores a variety of neural network architectures to classify ECG heartbeat signals effectively. 📌 Project Overview
We investigate and compare the performance of different neural network models for ECG signal classification, including:
Base architectures:
GRU (Gated Recurrent Unit)
LSTM (Long Short-Term Memory)
Attention-based models
Hybrid architectures that combine:
Convolutional Neural Networks (CNNs)
One of the base models above
We evaluate the strengths, weaknesses, and typical prediction errors of each model. Data augmentation techniques are also applied to assess their impact on model performance. 🎯 Research Objectives
This project seeks to answer the following key questions:
Which neural network architecture achieves the highest classification performance?
What are the differences in prediction errors between base and hybrid models?
What effect does data augmentation have on model performance?