Machine Learning Models Portfolio

Author: Enrico Miguel Veloso
Course: DSA#4155 - Artificial Intelligence
Institution: University of Santo Tomas - College of Science

---

Overview

This repository contains a comprehensive collection of machine learning implementations ranging from fundamental regression techniques to advanced ensemble methods and unsupervised learning. Each notebook demonstrates practical application of theoretical concepts using real-world datasets, with detailed explanations, visualizations, and performance evaluations.

---

Repository Structure

• Ordinary Least Squares Regression.md
• Linear Probability Model.md
• Logistic Regression and General Linear Models (GLM).md
• Regularization and Advance Classification.md
• Machine Learning Model Comparison.md
• Ensemble Methods and Advance Classifications.md
• Boosting Methods and Advance Ensemble Learning.md
• K-Means vs DBSCAN.md
• Dimensionality Reduction.md
• Market Basket Analysis.md

---

Models & Techniques Covered

Regression Models

Model	Dataset	Key Techniques
Ordinary Least Squares (OLS)	Auto MPG	Linear regression, VIF analysis, multicollinearity detection

Classification Models (Linear & Generalized)

Model	Dataset	Key Techniques
Linear Probability Model (LPM)	Adult Income (Census 1994)	Binary classification, probability bounds analysis
Logistic Regression	Adult Income (Census 1994)	GLM, odds ratios, statistical inference

Regularization Methods

Model	Techniques
Regularized Linear Models	Lasso (L1), Ridge (L2), Elastic Net, Cross-validation tuning

Tree-Based & Ensemble Methods

Model	Techniques
Decision Trees	Recursive partitioning, feature importance, pruning, visualization
Bagging	Bootstrap aggregation, variance reduction
Random Forest	Feature randomization, parallel ensemble, out-of-bag error
AdaBoost	Adaptive boosting, sample weighting, weak learners
Gradient Boosting	Residual fitting, sequential error correction

Unsupervised Learning

Model	Dataset	Techniques
K-Means Clustering	Credit Card Dataset	Elbow method, silhouette analysis, centroid interpretation
DBSCAN	Credit Card Dataset	Density-based clustering, eps & min_samples tuning, noise detection
PCA & Dimensionality Reduction	Global Country Data	Principal components, explained variance, biplots, t-SNE, UMAP

Association Rule Mining

Model	Dataset	Techniques
Market Basket Analysis	2022 Philippine Election Data	Apriori algorithm, support/confidence/lift metrics, rule mining

---

Dataset Descriptions

Dataset	Source	Size	Target Variable
Auto MPG	UCI ML Repository	398 rows, 9 features	Fuel efficiency (MPG)
Adult Income	OpenML (v2)	48,842 rows, 15 features	Income >$50K
Credit Card Dataset	Kaggle	8,636 rows, 6 features	Customer segments
Global Country Data 2023	Kaggle	195 rows, 10+ features	Development indicators
Philippines Election 2022	Figshare (CC BY 4.0)	Province-level	Senatorial voting patterns

---

🔧 Technologies Used

Core Libraries

importpandasaspdimportnumpyasnpimportmatplotlib.pyplotaspltimportseabornassns