Predicting Fixation Paths in the Moran Process:A Machine Learning Approach
Hajihashemi, M.
Show abstract
Path of Fixation in evolutionary process highly depends on structure of underlying population. In this paper, we apply a machine learning method to predict the path of fixation in several complex graphs and two regular graphs. In our approach, the path of fixation is not used as the target variable in the machine learning model. Rather, we focus on predicting the probability of progression forward (referred to as{lambda} in the literature) using the machine learning model. By using previous achievements in determining the fixation path for the Moran process, obtaining the path of fixation becomes straightforward. Due to the time and computational resources required for simulating an evolutionary process in a large population, utilizing a machine learning method can help us save both of these valuable resources. This approach can provide insights to researchers studying evolutionary processes in the context of meta-population problems.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.