Gilles Tonnelier The goal of machine learning is to design algorithms that can learn and generate statistical models for data analysis and prediction. Without being particularly trained for a given activity, the ML algorithms should be able to learn on their own and generate correct predictions depending on the data presented.

Thank you for your reply even if I'm not seeing the point with my question.

If y try to simplify a bit more my question, I'm looking for feedback concerning strategy from people doing twins models.

hope you the best.

Gilles Tonnelier ,

While I am not a biologist, I will try to answer your question

If you look at physiology from a historical perspective it is mostly descriptive of major components of the body (top down approach). As cell biology has progressed in the last decades a more bottoms up approach to physiology has taken place (i.e. systems biology).

Now the type of modelling depends on the particular approach that you assume. In systems biology some of the phenomena are modelled using differential equations while other more complex interactions are modelled using complex network theory. In summary, the body is still an extremely complex system that we are still trying to grapple with and much depends on what you are trying to specifically model.

Regards

Dear Gilles

Maybe I can contribute to your question. As my response to this is a bit longer please contact me under [email protected]. Beat