Flight Delays with Regression

Requirements Clarification & Assessment

1. Understanding the Problem Scope:

   • The primary goal is to forecast the likelihood of flight delays using a regression model.
   • The challenge arises from flights experiencing delays up to 12 hours, which skew the model's accuracy.

2. Data Examination:

   • Analyze the distribution of the delay times to confirm the presence and impact of outliers.
   • Assess the frequency and characteristics of these extreme delays to determine their influence on the model.

3. Model Objectives:

   • Decide whether the focus is on predicting the delay duration or simply identifying if a delay will occur.
   • Determine if the model should handle continuous delay times or categorize delays into distinct groups.

4. Stakeholder Requirements:

   • Confirm with stakeholders if they prefer a binary outcome (delay or no delay) or a more granular prediction (e.g., short, medium, long delay).