Experiments

The basic purpose of the experiments module of ai4water is comparison. It can be used for following scenarios

1. Comparison between different machine learning algorithms for a classification or regression task.

   This can be done using MLRegressionExperiments or MLClassificationExperiments class.

2. Comparison between different neural network architectures for a classification or a regression task.

   This can be done using DLRegressionExperiments or DLClassificationExperiments classes.

3. Test a single algorithm in different scenarios e.g., by applying different transformations on a feature and compare the results

   A use case is shown with TransformationExperiments class.

4. Optimize the hyperparameters of multiple models. This can be done by setting run_type to optimize

   in experimnnt.fit method.

All the classes inherit from Experiments class.

Machine learning algorithms for classification

Machine learning algorithms for classification

Neural Networks for classification

Neural Networks for classification

Comparison of machine learning algorithms

Comparison of machine learning algorithms

Comparison of deep learning architectures

Comparison of deep learning architectures

Comparison of LSTM with different transformations

Comparison of LSTM with different transformations

Comparison of XGBRegressor with different transformations

Comparison of XGBRegressor with different transformations