Shedrackeze002/Data-Inference-and-Applied-Machine-Learning

CMU-Africa: Data, Inference, and Applied Machine Learning course assignments # Data, Inference, and Applied Machine Learning Graduate coursework from **Carnegie Mellon University Africa** — MS in Engineering Artificial Intelligence. This repository contains assignments covering statistical inference, machine learning fundamentals, and applied data science. ## 📚 Assignments Overview ### Assignment 1: Exploratory Data Analysis & Statistical Foundations - **Topics:** Data loading, cleaning, visualization, descriptive statistics - **Datasets:** World Bank indicators (CO₂ emissions, GDP per capita, fertility rates, education metrics) - **Skills:** Pandas, Matplotlib, Seaborn, correlation analysis ### Assignment 2: Time Series Analysis — EirGrid Energy Data - **Topics:** Time series preprocessing, seasonality analysis, autocorrelation - **Dataset:** Ireland/Northern Ireland electricity grid data (15-min intervals, 2024) - **Key Tasks:** - Data quality checks and daylight saving time handling - Monthly, weekly, and daily aggregations - ACF analysis for wind and solar generation - Weekend vs. weekday demand patterns (hypothesis testing) - **Skills:** Time series decomposition, statistical testing, energy analytics ### Assignment 3: Classification — Loan Status Prediction - **Topics:** Binary classification, model evaluation, feature importance - **Dataset:** Loan application data with approval outcomes - **Models:** Logistic Regression, Decision Trees, ensemble methods - **Key Tasks:** - Feature engineering and outlier handling - ROC curves, confusion matrices, precision-recall analysis - SHAP values for model interpretability - **Skills:** Scikit-learn, SHAP, imbalanced classification ### Assignment 4: Regression — Life Expectancy Prediction - **Topics:** Regression analysis, feature selection, model comparison - **Dataset:** WHO Life Expectancy dataset - **Models:** Gradient Boosting Regressor, feature importance analysis - **Key Tasks:** - Missing value imputation strategies - Feature selection and dimensionality reduction - Model perfo …