Franz Wilhelmstötter

Do some experiments calculating the fitness function using the [Elo-rarting](https://en.wikipedia.org/wiki/Elo_rating_system). Find out if there are some use-cases for this kind of fitness function.

Implement the NSGA3 algorithm for MO optimization problems. * https://www.egr.msu.edu/~kdeb/papers/k2012009.pdf * https://www.cogentoa.com/article/10.1080/23311916.2016.1269383.pdf * https://yarpiz.com/456/ypea126-nsga3 * https://www.hindawi.com/journals/ijae/2018/8302324/

The `Random` engine, used by the library, is currently injected _statically_ with the help of the `RandomRegistry` class. This works good in the current environment. But with the Java project...

The currently used tooling for doing GA experimentation can be improved. Especially the data format for storing the results. The results are currently stored as XML file. This can be...

Add a Genetic Programming example, which tries to approximate a constant time series stream. The difference to the _Symbolic regression_ example is the, probably, infinite stream of input data.

The factory methods for `Chromosome`, `Genotype` and `Phenotype` are taking instances of `ISeq` objects. This can be relaxed to `BaseSeq` classes.