itm-unipi / ECG-Fitting-Forecasting-and-Activity-Classification

University Project for "Intelligent Systems" course (MSc Computer Engineering @ University of Pisa). ECG Fitting, Forecasting and Activity Classification using NN, CNN, RNN and Fuzzy Systems.

ECG Fitting, Forecasting and Activity Classification

University Project for "Intelligent Systems" course (MSc Computer Engineering @ University of Pisa). Implementation of ECG fitting, forecasting and activity classification using MLP and RBF neural networks, convolutional neural networks, recurrent neural networks and fuzzy systems.

Overview

The aim of the project is to apply machine learning, deep learning and fuzzy system technologies in order to classify human activities (such as walking, sitting and running), fit ECG feature values (such as mean and standard deviation) and forecast ECG values. To achieve these goals, the following technologies have been used:

• Multi-layer Perceptron, Radial Basis Function and Convolutional Neural Network for ECG feature values fitting
• Multi-layer Perceptron, Fuzzy Inference System and Adaptive Network-based Fuzzy Inference System for activity classification
• Recurrent Neural Network for ECG forecasting

Project Architecture

ECG-Fitting-Forecasting-and-Activity-Classification
├── docs
├── resources
└── src
    ├── data_preprocessing
    ├── neural_networks_ecg_fitting
    ├── neural_network_activity_classification
    ├── fuzzy_inference_system
    ├── convolutional_neural_network
    └── recurrent_neural_network

Authors

• Biagio Cornacchia
• Gianluca Gemini
• Matteo Abaterusso