Now you are on the step for Neural Network Modeling. You can choose the number of inputs upon your experience or you can get some idea from the correlation page. We are using the The nonlinear autoregressive network with exogenous inputs (NARX) method in our background for our time series modeling. You can read more about this alogrithm from Matlab. Considering a case for our paper. We will walk you through the process for the prediction of trajectory.
We have condiser inputs for our trajectory are Sway Velocity, Surge Velocity, Position East and Position North.
For output We have condiser Position East and Position North.
Now to start selecting your inputs , desired number of hidden neurons and outputs. Once you finished Press the "Start Modeling" Button.
Press control to multiple selection.

Matlab Will run background after you start Modeling.The Result is stored in JSON files. Now Click on Display results to see the Neural Network result. The plot will show Actual Trajectory vs Predicted Trajectory.