Exploring the depths of machine learning through hands-on projects, from supervised learning algorithms to unsupervised pattern discovery.
Powerful technologies used to build intelligent ML solutions
Explore implementations across supervised and unsupervised learning
Boston Housing Dataset
Implemented linear regression to predict house prices based on features like location, size, and amenities. Includes feature engineering, outlier detection, and model evaluation with R² and RMSE metrics.
Wisconsin Breast Cancer Dataset
Binary classification model to predict malignant vs benign breast cancer cases. Features comprehensive data preprocessing, feature selection, and model interpretation with confusion matrix analysis.
MNIST Dataset
Multi-class classification using KNN algorithm to recognize handwritten digits (0-9). Includes distance metric optimization, k-value tuning, and visualization of decision boundaries.
Enron Email Dataset
SVM implementation for binary text classification to identify spam emails. Features TF-IDF vectorization, kernel selection, hyperparameter tuning, and ROC curve analysis.
Loan Prediction Dataset
Decision tree classifier to predict loan approval based on applicant demographics and financial history. Includes tree visualization, feature importance analysis, and pruning techniques to prevent overfitting.
Credit Card Fraud Dataset
Advanced ensemble method using XGBoost for fraud detection in credit card transactions. Handles imbalanced data with SMOTE, implements cross-validation, and achieves high precision-recall performance.
Mall Customer Dataset
Unsupervised clustering algorithm to segment customers based on purchasing behavior and demographics. Includes elbow method for optimal k selection, cluster interpretation, and business insights visualization.
Reuters News Dataset
Agglomerative hierarchical clustering for grouping similar documents based on content. Features TF-IDF text preprocessing, dendrogram visualization, and optimal cluster number determination.