Machine Learning Projects 🎓

A curated collection of small-scale machine learning experiments completed during my undergraduate studies.
Each project focuses on classification models and exploratory data analysis (EDA), built using Python.
Projects are organized into self-contained scripts or Streamlit apps, featuring clean visualizations and straightforward logic for learning and demonstration purposes.

---

📁 Project Descriptions

🔹 wine-prediction

🔍 Features

A machine learning app that predicts wine quality based on chemical properties. It includes:

• EDA (correlation matrix, bar charts)
• Multiple ML models: Decision Tree, Random Forest, Logistic Regression
• Interactive Streamlit interface for model comparison and custom predictions

🛠 How to Run

cd wine-prediction
pip install -r requirements.txt
streamlit run wine_prediction.py

🔹 glass-type-prediction

A classification project that predicts the type of glass based on its chemical composition using multiple machine learning algorithms.

🔍 Features

• Exploratory Data Analysis (EDA):
  • Histograms, boxplots, scatter plots, and correlation heatmap
• Multiple classification models implemented:
  • Logistic Regression
  • Naive Bayes
  • K-Nearest Neighbors (KNN)
• Model evaluation using accuracy, prediction comparison, and confusion matrix
• Model persistence using joblib (KNN saved as .pkl)

🛠 How to Run

cd glass-type-prediction
pip install -r requirements.txt
python glass_prediction.py

🔹 cardio-prediction

A machine learning project that predicts the risk of cardiovascular disease based on health-related features such as blood pressure, cholesterol, BMI, and more.

🔍 Features

• Data preprocessing and EDA (age groupings, histograms, correlation heatmaps)
• Multiple classification models:
  • Logistic Regression
  • Decision Tree
  • Random Forest
  • K-Nearest Neighbors
  • Naive Bayes
• Model evaluation using accuracy scores and confusion matrix

🛠 How to Run

cd cardio-prediction
pip install -r requirements.txt
python cardio_prediction.py

🧠 Goals

• Practice data preprocessing, visualization, and model training
• Compare classifier performance in simple datasets
• Build interactive demos using Streamlit
• Keep code clean, modular, and reproducible

🔧 Tools Used

Python

Pandas, NumPy

Scikit-learn

Seaborn, Matplotlib

Streamlit

Joblib