euranova/aviation-data-challenge-20...

phases of the flight. We use a combination of a Kalman filter and a neural network to segment these phases. The Kalman filter helps smooth the data, while the neural network identifies the boundaries of the takeoff and climb phases. Once segmented, we extract relevant statistics from these phases to use as features in the model. ### Usage Run the script using: ```bash python feature_extractor/climb_takeoff_segmentation.py ``` The output is a set of features related to the climb and takeoff phases, which can be used for further analysis or modeling. ### Model Training and Evaluation The main script for training and evaluating the model is `main.py`. This script performs the following steps: 1. Reads the configuration file. 2. Downloads the dataset from MinIO. 3. Performs feature engineering using the specified method (general feature extraction or climb & takeoff segmentation). 4. Trains an XGBoost regression model with hyperparameter tuning using Optuna. 5. Evaluates the model using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). 6. Generates predictions for the submission set and saves them to a CSV file. 7. Visualizes feature importance. ### Usage Run the script using: ```bash python main.py ``` The script will