← Skill graph
Classical model evaluation
Part of Machine learning foundations
Needs first
Helps first
Nothing recommended beyond what is required.
Unlocks
Where it is assessed
14 assessed items, 0 done.
- AIE-107 M1.4 Three ways to the same answer · tier 1
- AIE-107 M2.4 Precision, recall, and what each error costs · tier 1
- AIE-107 M2.5 Is your confidence a probability? · tier 1
- AIE-107 M2.6 Retention: classification metrics and calibration · tier 1
- AIE-202 M2.3 Learning curves that actually diagnose · tier 1
- AIE-202 M2.4 Diagnose something you actually run · tier 2
- AIE-202 M2.5 Retention: reading a learning curve · tier 1
- AIE-202 M3.3 Find the leak · tier 2
- AIE-202 M4.3 Where clustering told you nothing · tier 2
- AIE-202 M5.3 A metric chosen for a reason · tier 1
- AIE-202 M5.4 Calibrate, and show the reliability diagram · tier 2
- AIE-202 M5.5 The baseline that should have been there · tier 3
- AIE-202 M5.6 Oral defense · tier 4
- AIE-202 M5.7 Retention: metric selection and calibration · tier 1
Verified sources
- Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow
- Machine Learning Crash Course
- A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification
- Selective Classification for Deep Neural Networks
- Probability calibration
- Metrics and scoring: quantifying the quality of predictions