QuestionQ87

VerifiedML – Data

An airline has developed an ML model to forecast fuel requirements for future flights. The model ingests weather data, including wind speeds and temperatures; determines flight routes from historical air-traffic-control routings; and estimates loads using average passenger and baggage weights. It met an acceptable airline standard throughout the summer, but once winter began, its load-weight estimates became less accurate. Exploratory data analysis showed that luggage weights were greater in winter than in summer.

Which statement BEST characterizes the problem and how it could have been avoided?

Explanation

The model’s input data distribution changed seasonally: winter baggage weights are higher than the summer values represented by its average-load assumptions. This is drift, and regular performance monitoring and testing can detect the degradation early so it can be mitigated, such as by updating the model or its inputs.

Community Discussion

No comments yet. Be the first to start the discussion!