Feature Engineering for Enhanced Flight Delay Prediction

[FreeSpirit11]

Description

To improve the accuracy and robustness of our flight delay prediction model, we need to create new features that may capture additional relevant information. The following new features should be added to the dataset:

1. Time of Day: Create a categorical feature indicating the time of day (morning, afternoon, evening, night) based on the scheduled departure time.
2. Day of the Week: Extract the day of the week from the date and add it as a new feature.
3. Weekend Indicator: Add a binary feature indicating whether the flight is scheduled on a weekend.
4. Distance: Calculate the distance between the origin and destination airports.
5. Historical Delay Patterns: Incorporate historical delay patterns for specific routes or airlines, potentially using past data to create features representing average delays for given routes or airlines.

Tasks

1. Create Time of Day Feature:
   • Categorize SCHEDULED_DEPARTURE into morning, afternoon, evening, and night.
2. Add Day of the Week Feature:
   • Extract the day of the week from the DAY and MONTH columns.
3. Weekend Indicator:
   • Create a binary feature to indicate whether the flight is on a weekend.
4. Calculate Distance:
   • Use geographical coordinates of the airports to calculate the distance between the origin and destination.
5. Historical Delay Patterns:
   • Analyze past delay data to create features representing average delays for specific routes or airlines.

Benefits

Adding these features is expected to provide the model with more context and improve its ability to predict flight delays accurately.

[FreeSpirit11]

Hi @Niketkumardheeryan , I have raised this issue . Please assign it to me.

[invigorzz313]

This issue has been assigned. @FreeSpirit11 Do follow the project guidelines for PR

[FreeSpirit11]

@invigorzz313 Please assign level 2 to this Issue then I will start working on it.