Heart Disease Detection from ECG Signals — ML Pipeline

This project implements a full ML workflow for binary classification of ECG signals (normal vs abnormal).

Key components:

Dataset: CSV file where each row is an ECG time-series sample and the last column is the label (0 = Normal, 1 = Abnormal).

Preprocessing: Extract features from raw time-series, handle noise, split into features (X) and labels (y).

EDA: Visualize random ECG traces, check class balance and signal patterns.

Models implemented:

Logistic Regression (baseline)

Support Vector Machine (SVM)

Naive Bayes

Evaluation: Accuracy, Precision, Recall, F1-score and classification report are computed for each model.

Outputs & Use-cases: Trained models and evaluation results can be extended for real-time monitoring, decision support tools, or deployment on embedded devices (IoT/MCU).

Files included: code notebook with preprocessing, plots, model training and evaluation. See attached file links below.