Machine Learning Model Evaluation

This project involves analyzing different machine learning models for classifying falls and activities of daily living (ADL). The process is divided into three main steps:

1. Dataset Creation (create_dataset.py)
2. Data Processing and Model Training (process.py)
3. Model Accuracy Visualization (plot_accurs.py)

Project Structure

• create_dataset.py: This script extracts features from raw data and creates training and test datasets.
• process.py: This script trains various machine learning models and evaluates their accuracies, saving the results to a CSV file.
• plot_accurs.py: This script loads the saved model accuracies and plots a bar chart showing the comparison between models.

Usage Instructions

Step 1: Dataset Creation

Run the create_dataset.py script to create the dataset with extracted features and prepare the training and test sets.

python create_dataset.py

Step 2: Data Processing and Model Training

After creating the dataset, run the process.py script to train the models and evaluate their accuracies. This script will save the results to a CSV file.

python process.py

Step 3: Model Accuracy Visualization

Finally, run the plot_accurs.py script to visualize the accuracy of the models in a bar chart. The chart will be saved as a PNG image and displayed on the screen.

python plot_accurs.py

Script Descriptions

create_dataset.py

This script performs the following operations:

1. Data Loading: Loads raw data from the Fall and ADL folders.
2. Feature Extraction: Extracts relevant features from the data.
3. Dataset Creation: Saves the training and test datasets to CSV files.

process.py

This script performs the following operations:

1. Data Loading: Loads the training and test datasets.
2. Model Training: Trains various machine learning models, including Perceptron, ADALINE, Logistic Regression, SVM, Decision Trees, Random Forests, k-Nearest Neighbors, AdaBoost, Gradient Boosting, XGBoost, and LightGBM.
3. Accuracy Evaluation: Evaluates the models and saves the accuracies to a CSV file.

plot_accurs.py

This script performs the following operations:

1. Accuracy Loading: Loads model accuracies from the CSV file.
2. Chart Generation: Plots a bar chart showing the accuracy of each model and saves the image as a PNG file.

Dependencies

Ensure you have the following libraries installed:

• pandas
• numpy
• matplotlib
• seaborn
• scikit-learn
• xgboost
• lightgbm

You can install these libraries using pip:

pip install pandas numpy matplotlib seaborn scikit-learn xgboost lightgbm

Contribution

Feel free to contribute improvements, bug fixes, or suggestions. To contribute, please submit a pull request or open an issue in the repository.