This repository contains two Python machine learning projects demonstrating data exploration, visualization, and classification. Both projects are fully reproducible and include their datasets, making them ideal for showcasing ML workflow, data analysis skills, and predictive modeling.
- Dataset:
bank_notes.csv - Goal: Predict whether a bank note is authentic or fake using wavelet features and image entropy.
- Features: Wavelet Variance, Wavelet Skewness, Wavelet Kurtosis, Image Entropy
- Methods: Logistic Regression, Linear Discriminant Analysis, K-Nearest Neighbors, Decision Trees, Naive Bayes, Support Vector Machines
- Highlights:
- Data visualization: box plots, histograms, scatter matrix
- Algorithm comparison with cross-validation
- Prediction on new bank notes
- Dataset:
student_stress_factors.csv - Goal: Predict student stress levels based on sleep quality, headaches, performance, and study load.
- Features: Sleep Quality, Headaches, Performance, Study Load
- Methods: Decision Tree Classifier
- Highlights:
- Data exploration: summary statistics, box plots, histograms, scatter matrix
- Fully reproducible predictions with fixed random state
- Predict stress levels for new student data
- Clone the repository:
git clone https://github.com/jennabeachcodes/Machine-Learning
- Install required packages:
pip install pandas matplotlib scikit-learn
- Run the Python scripts for each project:
python bank_notes_ml.py python student_stress_ml.py
- Data loading and cleaning
- Univariate and bivariate analysis
- Data visualization (box plots, histograms, scatter matrices)
- Supervised machine learning (classification)
- Model evaluation and prediction
- Reproducible code for portfolio-quality projects
This project is provided for educational purposes only as part of the course 26W-CST8400 - Analysis and Design Using Emerging Technologies at Algonquin College, Ottawa, ON, Canada.