Machine Learning Projects

This repository contains two main machine learning projects:

1. Adult Census Income Prediction

Objective: Predict whether an individual's income exceeds $50K/year based on census data.

Dataset:

• UCI Adult Census dataset (CSV format)
• Features include age, workclass, education, marital status, occupation, race, sex, capital gain/loss, hours per week, native country, and salary class.

Process:

• Data cleaning and preprocessing (handling missing values, encoding categorical variables, feature engineering)
• Exploratory Data Analysis (EDA): Univariate and bivariate analysis of features
• Model building: Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree, Random Forest, Neural Network
• Model evaluation: Accuracy, F1-score, confusion matrix, classification report
• Ensemble modeling for improved performance

Results:

• Multiple models compared; ensemble model achieves the best accuracy and F1-score on both train and test sets.

2. Churn Prediction

Objective: Predict customer churn for a financial institution using customer transaction and demographic data.

Dataset:

• Custom churn dataset (churn_prediction.csv)
• Features include customer demographics, account balances, transaction history, dependents, occupation, city, branch code, etc.

Process:

• Data cleaning (handling missing values, outlier treatment, feature transformation)
• Feature engineering and encoding (label and one-hot encoding)
• Exploratory Data Analysis (EDA): Univariate and bivariate analysis
• Model building: Logistic Regression, KNN, Decision Tree, Random Forest
• Model evaluation: Accuracy, F1-score
• Ensemble modeling for final predictions

Results:

• Ensemble model provides robust predictions and improved accuracy compared to individual models.

---

How to Run

1. Open the Jupyter notebooks (Adult_census_income.ipynb and churn_prediction/churn_prediction.ipynb) for step-by-step code and analysis.
2. Ensure required Python packages are installed: numpy, pandas, matplotlib, seaborn, scikit-learn, tensorflow (for neural network).
3. Place the datasets in the correct paths as referenced in the notebooks.

Project Structure

• Adult_census_income.ipynb: Notebook for census income prediction
• churn_prediction/churn_prediction.ipynb: Notebook for churn prediction
• churn_prediction/churn_prediction.csv: Dataset for churn prediction
• README.md: Project overview and instructions

---

For details, see the notebooks for code, analysis, and results.