Predictive Analytics

  1. Model Evaluation and Validation
    1. Data Splitting Strategies
      1. Hold-out Validation
        1. Training Set Size
          1. Validation Set Purpose
            1. Test Set Independence
            2. Time-based Splitting
              1. Temporal Validation
                1. Walk-forward Analysis
                  1. Expanding Window
                    1. Rolling Window
                    2. Stratified Sampling
                      1. Class Balance Preservation
                        1. Stratification Variables
                      2. Cross-Validation Techniques
                        1. K-Fold Cross-Validation
                          1. Fold Selection
                            1. Variance Estimation
                              1. Computational Considerations
                              2. Stratified K-Fold
                                1. Class Distribution Maintenance
                                  1. Imbalanced Data Handling
                                  2. Leave-One-Out Cross-Validation
                                    1. Bias-Variance Properties
                                      1. Computational Complexity
                                      2. Time Series Cross-Validation
                                        1. Temporal Dependencies
                                          1. Forecast Horizon Considerations
                                          2. Nested Cross-Validation
                                            1. Model Selection and Assessment
                                              1. Hyperparameter Tuning
                                            2. Regression Evaluation Metrics
                                              1. Error-based Metrics
                                                1. Mean Absolute Error
                                                  1. Mean Squared Error
                                                    1. Root Mean Squared Error
                                                      1. Mean Absolute Percentage Error
                                                      2. Relative Metrics
                                                        1. R-squared
                                                          1. Adjusted R-squared
                                                            1. Mean Absolute Scaled Error
                                                            2. Distribution-based Metrics
                                                              1. Quantile Loss
                                                                1. Pinball Loss
                                                              2. Classification Evaluation Metrics
                                                                1. Confusion Matrix Analysis
                                                                  1. True Positives and Negatives
                                                                    1. False Positives and Negatives
                                                                      1. Error Types
                                                                      2. Single-value Metrics
                                                                        1. Accuracy
                                                                          1. Precision
                                                                            1. Recall
                                                                              1. F1-Score
                                                                                1. Specificity
                                                                                2. Threshold-dependent Analysis
                                                                                  1. ROC Curve
                                                                                    1. Precision-Recall Curve
                                                                                      1. Area Under Curve
                                                                                      2. Multi-class Extensions
                                                                                        1. Macro Averaging
                                                                                          1. Micro Averaging
                                                                                            1. Weighted Averaging
                                                                                            2. Class Imbalance Considerations
                                                                                              1. Balanced Accuracy
                                                                                                1. Matthews Correlation Coefficient
                                                                                                  1. Cohen's Kappa
                                                                                                2. Model Comparison and Selection
                                                                                                  1. Statistical Significance Testing
                                                                                                    1. Paired t-test
                                                                                                      1. McNemar's Test
                                                                                                        1. Wilcoxon Signed-rank Test
                                                                                                        2. Information Criteria
                                                                                                          1. Akaike Information Criterion
                                                                                                            1. Bayesian Information Criterion
                                                                                                              1. Cross-validation Information Criterion
                                                                                                              2. Hyperparameter Optimization
                                                                                                                1. Grid Search
                                                                                                                  1. Random Search
                                                                                                                    1. Bayesian Optimization
                                                                                                                      1. Evolutionary Algorithms
                                                                                                                    2. Bias-Variance Analysis
                                                                                                                      1. Bias-Variance Decomposition
                                                                                                                        1. Overfitting Detection
                                                                                                                          1. Learning Curves
                                                                                                                            1. Validation Curves
                                                                                                                              1. Complexity Analysis
                                                                                                                              2. Underfitting Identification
                                                                                                                                1. Model Capacity Assessment
                                                                                                                                  1. Feature Adequacy
                                                                                                                                  2. Regularization Effects
                                                                                                                                    1. Regularization Paths
                                                                                                                                      1. Cross-validation for Regularization

                                                                                                                                  Previous

                                                                                                                                  8. Unsupervised Learning

                                                                                                                                  Go to top

                                                                                                                                  Next

                                                                                                                                  10. Model Interpretability and Explainability