InteractiveMap of Operations Research
Back to Modern OR Practice
TheoryAdvanced06.01
Modern OR Practice

Analytics, Data Science & ML

Predict, then prescribe.

Overview

Analytics, Data Science & ML focuses on predict, then prescribe. In the map of OR, it connects Forecasting, Predict-then-optimize, Decision-focused learning to decisions that must be modeled, solved, explained, and revised as evidence changes.

Modern OR combines forecasting and machine learning with optimization: decision-focused learning, predict-then-optimize, bandits, and RL. The practical use case is clearest in adjacent OR applications, where the method helps turn constraints and tradeoffs into a decision artifact someone can inspect.

Core ideas

Forecasting

Forecasting is a core checkpoint for Analytics, Data Science & ML: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

Predict-then-optimize

Predict-then-optimize is a core checkpoint for Analytics, Data Science & ML: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

Decision-focused learning

Decision-focused learning is a core checkpoint for Analytics, Data Science & ML: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

Bandits

Bandits is a core checkpoint for Analytics, Data Science & ML: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

RL

RL is a core checkpoint for Analytics, Data Science & ML: define it concretely, attach units or rules where possible, and test whether stakeholders interpret it the same way.

How to use it

  1. 1Start with a concrete case from the surrounding OR area: write the decision, time horizon, actors, and objective in operational language.
  2. 2Translate the problem into Forecasting, Predict-then-optimize, and Decision-focused learning; define units and data sources for each one.
  3. 3Build a small instance of Analytics, Data Science & ML that can be solved or simulated by hand inspection before using full production data.
  4. 4Compare the recommendation against a baseline policy, not just against mathematical optimality.
  5. 5Document assumptions, sensitivity results, and the conditions under which the recommendation should be revisited.

Applications

Use this topic as a building block in nearby OR models; connect it to a concrete decision before treating it as a standalone application area.

Common pitfalls

  • Applying Analytics, Data Science & ML because the label sounds appropriate while leaving the actual decision boundary vague.
  • Treating Forecasting as a technical detail instead of a modeling choice that affects the recommendation.
  • Reporting one answer without showing sensitivity to demand, capacity, costs, or behavioral assumptions.
  • Ignoring implementation details such as data quality, explainability, ownership, and how users will override bad recommendations.

Resources