Performance Metrics for a Classification Model
Core Metrics DS practice problem on Onlearn.
Difficulty: medium.
Topics: Performance Metrics for a Classification Model, F1-Score, Area Under the ROC Curve, Logarithmic Loss, Precision-Recall Tradeoff, Matthews Correlation Coefficient, Supervised Learning, Statistical Inference, Information Theory, Decision Theory, Data Quality Assurance, Binary Classification, Confusion Matrix Analysis, Probabilistic Thresholding, Cost-Sensitive Learning, Model Calibration.
Task: Implement Performance Metrics Calculation In this task, you are required to implement a function performance metrics(actual, predicted) that computes various performance metrics for a binary classification problem. These metrics include: Confusion Matrix Accuracy F1 Score Specificity Negative Predictive Value The function should take in two lists: actual: The actual class labels (1 for positive, 0 for negative). predicted: The predicted class labels from the model. Output The function should return a tuple containing: 1. confusion matrix: A 2x2 matrix. 2. accuracy: A float representing the accuracy of the model. 3. f1 score: A float representing the F1 score of the model. 4. specificity: A float representing the specificity of the model. 5. negative predictive value: A float representing the negative predictive value. Constraints All elements in the actual and predicted lists must be either 0 or 1. Both lists must have the same length.