LAUSR

Significance Evolution is a historical process with contingency, where outcomes are sensitive to past events. Due to this, predicting evolution remains challenging. In this paper, we propose a method to predict the response to selection that incorporates history. The method uses tools from quantitative genetics and combines them with information from past evolution that is extracted through a combination of signal processing and learning techniques. This information is underexploited by existing methods to predict evolution but is of great value since it reflects singularities of the evolutionary system. We show that this combination of information coming from the time series and quantitative genetics methods outperforms classical methods in predicting the response to selection.