Model Implementation and Training

Implement your chosen anomaly detection model, integrating `Featuretools` for feature generation. Train the model using the provided simulated telemetry data. Pay close attention to handling imbalanced datasets (anomalies are rare) and optimizing model hyperparameters. Detail how `Gemini 2.5 Flash` is integrated into your model architecture for identifying complex patterns. Document your training process, including data split, validation techniques, and the performance metrics you used to select your best model.