In this project, seasonal species migration is modeled. Many species show bimodal migratory pattern depicting spring and autumn migrations
Build a neural network to estimate a non-linear function that models the bimodal migratory pattern.
Fig.1 Network prediction superimposed on data. Data source: Samarasinghe S.[1]. Graphic made by my own.
Blue points correspond to the original data for migration. The orange line is the predicted non-linear function by the neural network with an 1-4-1 architecture, 20,000 epochs and a learning rate of 1.
[1] Samarasinghe, S. Neural Networks for Applied Sciences and Engineering. From Fundamentals to Complex Pattern Recognition, Auerbach Publications, pp.94-97, 2007.