Section 1 of 4 · 3 lessons

Define the ML Problem

A model can produce a prediction without helping anyone make a better decision. We will start with Botie's inspection team, work out what they can act on, and decide what a learned model would need to improve. The information available at decision time will become part of the design.

0 of 3 finished

Decisions Before Models

A prediction helps only if someone can use it. Begin with the inspection decision, identify its limits, and compare a learned model with a rule we can already run.

  1. 01What Would This Prediction Actually Change?A model can flag a pump that may fail, but someone still has to decide what to do. Begin with that decision and work backwards to the information it needs.
    Build with Botie →Read the lesson
  2. 02Do We Need a Learned Model Yet?A fixed rule already catches some faults. Work out what a learned model would have to improve before paying for its extra work.
    Build with Botie →Read the lesson

The Information We Can Use

A record in today's database may have arrived too late for yesterday's decision. Define what each input means and when the service could actually use it.

  1. 03What Can the System Know at Decision Time?A reading can describe an earlier event and still arrive too late. Design a data contract that respects both clocks.
    Build with Botie →Read the lesson