logikbot

Course Introduction

Course Outcomes

Course Structure

Imbalanced Classification Defined

Causes of Class Imbalance

Challenge of Imbalance Classification

Examples of Class Imbalance

Create Synthetic Dataset with Class Distribution

Effect of Skewed Class Distributions

Visualizing Extreme Skew

Why Imbalanced Classification Is Hard

Compounding Effect of Dataset Size

Compounding Effect of Label Noise

Compounding Effect of Data Distribution

Evaluation Metrics and Imbalance

Taxonomy of Classifier Evaluation Metrics

Ranking Metrics for Imbalanced Classification

Probabilistic Metrics for Imbalanced Classification

How to Choose an Evaluation Metric

Accuracy Fails for Imbalanced Classification

Accuracy Paradox

Demo: Accuracy for Imbalanced Classification

Precision for Imbalanced Classification

Precision for Multi-Class Classification

Recall for Imbalanced Classification

Demo: Recall for Imbalanced Classification

F-Measure for Imbalanced Classification

Demo: F- Measure for Imbalanced Classification

ROC Curves and Precision-Recall Curves

ROC Curve

Demo: ROC Curve

ROC Area Under Curve (AUC) Score

Precision-Recall Curves

Precision-Recall Area Under Curve (AUC) Score

ROC AUC on with Severe Imbalance

ROC and Precision-Recall Curves With a Severe Imbalance

Probability Scoring Methods in Python

Log Loss Score

Brier Score

Cross-Validation for Imbalanced Classification

Challenge of Evaluating Classifiers

Failure of k-Fold Cross-Validation

Data Sampling Methods for Imbalanced Classification

Oversampling Techniques

Undersampling Techniques

Combinations of Techniques

Random Resampling Imbalanced Datasets

Demo: Random Oversampling Imbalanced Datasets

Demo: Random Undersampling Imbalanced Datasets

Demo: Combining Random Oversampling and Undersampling Techniques

Synthetic Minority Oversampling Technique (SMOTE)

SMOTE for Balancing Data

SMOTE for Classification

Borderline-SMOTE SVM

Adaptive Synthetic Sampling (ADASYN)

Undersampling Methods

Near Miss Undersampling (NearMiss-1)

Near Miss Undersampling (NearMiss-2 and NearMiss-3)

Condensed Nearest Neighbor Rule Undersampling

Tomek Links for Undersampling

Edited Nearest Neighbors Rule for Undersampling (ENN)

Neighborhood Cleaning Rule for Undersampling

Cost-Sensitive Learning for Imbalanced Classification

Not All Classification Errors Are Equal

Cost-Sensitive Learning

Cost-Sensitive Imbalanced Classification

Cost-Sensitive Methods

Cost-Sensitive Algorithms

Cost-Sensitive Ensembles

Cost-Sensitive Logistic Regression

Logistic Regression for Imbalanced Classification

Weighted Logistic Regression with Scikit-Learn

Grid Search Weighted Logistic Regression

Cost-Sensitive Decision Trees for Imbalanced Classification

Decision Trees for Imbalanced Classification

Weighted Decision Tree With Scikit-Learn

Grid Search Weighted Decision Tree

Develop a Cost-Sensitive Neural Network for Imbalanced Classification

Neural Network Model in Keras

Deep Learning for Imbalanced Classification

Weighted Neural Network With Keras

Project: Breast Cancer Dataset

Haberman Breast Cancer Survival Dataset

Dataset Exploration

Model Test and Baseline Result

Evaluate Probabilistic Models

Model Evaluation With Scaled Inputs

Model Evaluation With Power Transform