further devopment of fw models

Description:

This issue outlines the steps and enhancements required for the development of forward models in Scilab

Tasks:

Ensembles

• Create a new class that maintains an array (or similar structure) of the parent class for ensemble-based modeling.

Pre-training Forward Models

• Implement pre-training with a parameter call.
• Incorporate learning from offline data to initialize models effectively before online training.

Overfitting Prevention

• Design train/test splits to minimize the risk of overfitting during model training.

One-step Prediction

• Implement one-step prediction functionality to facilitate model predictive control.