Human Activity Recognition using Smartphones

Overview

This project implements Human Activity Recognition (HAR) using the UCI HAR Dataset. It compares the performance of a Gaussian Naive Bayes classifier trained on:

• All available features
• Features reduced using K-Means clustering

The dataset is downloaded automatically from the UCI Machine Learning Repository.

---

Dataset

Human Activity Recognition Using Smartphones Dataset

• Source: UCI Machine Learning Repository

---

Technologies Used

• Python
• Pandas
• NumPy
• Scikit-learn
• BeautifulSoup
• Requests

---

Machine Learning Workflow

1. Download dataset automatically
2. Load training data
3. Encode activity labels
4. Standardize features
5. Train Gaussian Naive Bayes on all features
6. Reduce features using K-Means clustering
7. Train Gaussian Naive Bayes on reduced features
8. Compare accuracy and training time

---

Results

The notebook prints:

• Accuracy using all features
• Accuracy using reduced features
• Training time
• Number of selected features

---

How to Run

pip install -r requirements.txt

Open the notebook and run all cells.

The dataset is downloaded automatically.