Syllabus Map
- Study map: Syllabus Study Map
Overview
- Data science fundamentals cover evaluation, validation, feature workflows, and clean data pipelines.
- Good practice separates data understanding, modelling, and monitoring into repeatable steps.
- Common failure modes come from leakage, biased data collection, or misaligned metrics.
Core Areas
Evaluation Metrics
- Choose metrics that match the business and error-cost objective.
- Use both aggregate metrics and diagnostic views (for example confusion matrix).
- Detailed note: Evaluation Metrics
Validation Strategy
- Keep train/validation/test logic strict to avoid leakage.
- Use cross-validation when data is limited or split variance is high.
- Detailed note: Validation Strategy
Feature Engineering
- Encode categorical variables safely and select features that reflect signal.
- Build statistical and temporal features carefully to avoid leakage.
- Detailed note: Feature Engineering
Data Processing
- Handle missing values, scaling, and augmentation with training-only fitting.
- Pipeline discipline is key for reproducibility and honest validation.
- Detailed note: Data Processing