Development of a wearable fall detection system using Machine Learning

This project was developed for the course of "Statistical Learning for Healthcare Data" held at Politecnico di Milano in the academic year 2022-2023.

Development of a wearable fall detection system utilizing machine learning algorithms. Analysis of accelerometer time series using advanced algorithms to identify patterns and distinguish between everyday activities and fall events.

---

About The Project

We present the development of a classification system able to distinguish several distinct activities of daily living based on accelerometer measurements. The dataset comprises 468 events, each representing an activity performed by a single participant. Each event includes the activity label and three time series measurements corresponding to the accelerometer readings on the X, Y, and Z axis.

Structure

• SLHD-report Project final report
• SLHD-slides.pdf Project final presentation
• 3-classes.ipynb Data pre-processing and development of the classification models for 3 classes

Prerequisites

Install packages in requirements.txt

scipy
numpy
seaborn
sklearn
matplotlib

Authors

• Filippo Cipriani
• Sofia Sonnacchi
• Andrea Puricelli

(back to top)