AI/ML 101 Course – Part 1 (12/10/2024)
Duration: 2 hours
Topic 1: Data Balancing
Instructor: Nathan Sahelijo
Presentation link:
- Introducing data balance.
- What do we mean by data balance
- How does data imbalance affect ML performance
- Introduce types of imbalance
- How to correct imbalance
- Introduce and implement over and under sampling methods
- Introduce and implement weighted loss correction
- Synthetic sampling using SMOTE
Topic 2: Regression Analysis and Loss Functions
Instructor: Nathan Sahelijo
Presentation link:
- Introduce basics of regression.
- How does linear and logistic regression work?
- What is a loss function?
- How can we correct for overfitting?
- How we implement a functional regression model.
- Introduce the idea of loss functions
- Introduce regularization methods
Topic 3: Decision Trees
Instructor: Dhawal Priyadarshi
Presentation link:
- Introduction to Decision Trees
- Tree structure (nodes/leaves) - model if-then statements
- Classification Trees vs Regression Trees
- CART algorithm
- Splitting criteria (tree growing)
- Pruning
- Improving supervised modeling
- Intro to Bias-variance/overfitting-underfitting
- Training - Cross-validation (very briefly)
- Ensembling - Introduction to Random Forests (very briefly) - ensembling weaker trees, explainability
Topic 4: Supervised Evaluation Metrics
Instructor: Dhawal Priyadarshi/Nathan Sahelijo
Presentation link:
- Classification Evaluation Metrics (Dhawal)
- Confusion Matrix
- Accuracy
- Precision/Recall
- F1
- AUC/ROC curves
- Regression Evaluation Metrics (Nathan)
- MAE/MAPE/MSE
- RMSE/RMSLE
- R2/adj R2
Topic 5: Clustering Analysis
Instructor: Ankit Kumar Mishra
Presentation link:
- Introduction to Clustering
- What is clustering
- How is it different from classification
- Types of clustering
- Distance and evaluation metrics
- How to perform clustering using scikit-learn
- Agglomerative Clustering
- K-means
- DBSCAN
Topic 6: Dimensionality Reduction
Instructor: Ankit Kumar Mishra
Presentation link:
- Introduction to Dimensionality Reduction
- What is Dimensionality Reduction
- Feature Extraction vs Feature Selection
- Types of Dimensionality Reduction
- Algorithms
- PCA
- t-SNE
- UMAP