Description

Prepare the labeled dataset for machine learning by cleaning, formatting, and structuring it for training. This step ensures high-quality input data for the model. All outputs must be stored inside the inference/ folder.

Tasks

• Create script prepare_data.py inside inference/
• Read dataset from inference/labeled_flight_data.csv
• Remove unnecessary columns (e.g., timestamp if not used for training)
• Handle missing or null values (drop or fill appropriately)
• Encode event_label into numerical format (e.g label encoding)
• Select input features: altitude, heading, vertical_speed, velocity, roll, pitch, yaw, g_force
• Separate features (X) and labels (y)
• Split dataset into training and testing sets (e.g 80/20 split)
• Normalize or scale feature values
• Save and push the processed dataset to inference/dataset
• Push code to inference/

Acceptance Criteria

• Dataset is cleaned and ready for ML training
• Features and labels are correctly separated
• Train-test split is implemented
• Encoded labels are consistent and valid