A comprehensive repository containing End-to-End Machine Learning and AI models trained across diverse CSV datasets using Supervised and Unsupervised learning architectures.
Welcome to my Machine Learning and Artificial Intelligence repository. This project serves as a comprehensive workspace tracking the implementation, training, evaluation, and optimization of various predictive and analytical models across diverse real-world CSV datasets.
- Regression Layouts: Linear, Multiple, Ridge/Lasso, and Polynomial models for continuous metric forecasting.
- Classification Trees: Logistic Regression, Support Vector Machines (SVM), Random Forests, Decision Trees, and Naive Bayes for discrete target sorting.
- Performance Tracking: Validated using Confusion Matrices, Precision-Recall curves, F1-Score, and ROC-AUC limits.
- Clustering Models: K-Means, Hierarchical Clustering (Dendrograms), and DBSCAN to segment hidden data patterns without labels.
- Dimensionality Reduction: Principal Component Analysis (PCA) to compress high-dimensional feature spaces.
- Data Sourcing: Multi-domain CSV datasets processed sequentially.
-
Feature Engineering: Custom pipelines handling missing value imputation, out-of-bounds capping, categorical tracking, and scale normalization (
StandardScaler/MinMaxScaler). -
Model Lifecycle: Exploratory Data Analysis (EDA)
$\rightarrow$ Preprocessing$\rightarrow$ Training$\rightarrow$ Hyperparameter Tuning (GridSearchCV)$\rightarrow$ Evaluation.
- Languages: Python (Core)
- Libraries: Pandas, NumPy, Scikit-Learn, SciPy
- Visualization: Matplotlib, Seaborn
- Format: Jupyter Notebooks & Modular
.pyscripts