Coursework for Advanced Machine Learning (prev. Advanced Predictive Modeling) at UT Austin
Following is the range of concepts and topics covered in the assignments:
- AML_HW01_RuchiS.ipynb - Adoption of AI; Maximum Likelihood Estimate; Linear Regression with Lasso, Ridge and Elastic Net regularization
- AML_HW02_RuchiS.ipynb - Bias Variance TradeOff; Learning Curves; Stochastic Gradient Descent and Momentum
- AML_HW03_RuchiS.ipynb - Tensorflow Playground, t-SNE & PCA Dimensionality Reduction, PyOD Outlier Detection (KNN, INNE), Preprocessing and Sampling
- AML_HW04_RuchiS.ipynb - Expected Loss for Classficiation; Logistic Regression expression and formulation; Classfication with Decision Tree, Logistic Regression, MLP and imbalanced classes handling using SMOTE; Comparison of ROC-AUC and Precision-Recall Curves; Bayes Decision Theory; Bayesian Belief Networks; Theoretical understanding of Support Vector Machines including Slack and Kernels
- AML_HW05_RuchiS.ipynb - Ensemble Methods, Comparison between CatBoost and XGBoost; Experimenting with Decision Tree, Bagging, Random Forest, AdaBoost, XGBoost; XGBoost on datasets with varying class imbalance and plotting calibration curves; Motivation behind Skip Connection and Batch Normalization in Deep Learning