AI and Digital Transformation

Machine Learning Foundations for Business Analysts

An analyst does not need to build models to be the person who decides whether one should be trusted, and that judgement rests on understanding what the model was trained on and what it was asked to predict. This course builds that judgement. Delegates work through supervised and unsupervised learning in plain terms, framing a business question as a prediction problem, and the data required to answer it. Training, validation and the specific way models fool their builders are covered. Evaluation metrics follow, including why accuracy is misleading when outcomes are rare. Overfitting, drift and the reasons a model that worked stops working are addressed. Bias and fairness are covered concretely rather than abstractly. Interpreting model output for a decision maker follows. The course closes on working with data scientists and on knowing when the answer is not a model.