Machine Learning (ML) · Intermediate level

Intermediate level · Module 5 of 6 · 14 guided hours

Imbalanced and Messy Real-World Data

Handle rare outcomes, missing data and changing conditions that real projects always bring.

Learning objectives

  • Handle class imbalance with weighting, resampling and thresholds
  • Choose imputation strategies and model missingness
  • Recognise data drift and plan for it

Topics

  • class imbalance
  • class_weight and resampling
  • SimpleImputer and missing indicators
  • outliers and errors
  • data drift and concept drift
  • validation that mirrors deployment

Lessons

  1. Lesson 1 Class Imbalance

    Build useful models when the event of interest is rare.

    imbalanceclass_weight='balanced'oversampling and undersamplingprecision–recall evaluation

  2. Lesson 2 Missing Data Strategies

    Choose and compare imputation strategies, and use the fact that data is missing as information.

    SimpleImputerKNNImputer and IterativeImputermissing indicatorwhy data is missing

  3. Lesson 3 Data Drift and Realistic Validation

    Recognise how changing conditions degrade models, and validate in a way that mirrors real use.

    data driftconcept drifttime-based splitmonitoring signals

Practical activity

Improve the cash-shortage model for rare events, compare imputation strategies and simulate a shift in agent behaviour to see how performance changes.

Expected outcomes

  • Models that cope with real data
  • Awareness of drift before deployment

Assessment

Practical comparison of imbalance and imputation strategies with a short drift analysis.

Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.