[Feat] Benchmarking tools

[bcebere]

Feature Description

The library should offer methods for evaluating predictive models/pipelines.

For each problem type, there can be different relevant metrics, as described in the linked tasks.

The evaluation should be done using KFold(regression)/StratifiedKFold(classification, survival analsysis), and predefined random seed.

AP reference: https://github.com/vanderschaarlab/autoprognosis/blob/main/src/autoprognosis/utils/tester.py
Example implementation for time series survival
https://github.com/vanderschaarlab/synthcity/blob/main/src/synthcity/plugins/core/models/time_series_survival/benchmarks.py#L142

blocked by #9
blocked by #10
blocked by #11