Solution of optimization problems
- Choose the number of input units to intialize the SOM. In this case, the input vector represents the coordinates of a city. So the input dimension is 2.
- Choose the number K of nodes/neurons in the output layer. If N denotes the number of cities, then K should be greater than or equal to N.
- Choose over the total number of iterations the SOM will go through. Each iteration involves adjustment of weights for all the neurons.
- Choose the step size for iteration results display purpose for number of cities i.e. units in the output network. Also choose the iteration step size for generation of output.
- Click on 'Next city' button or the 'Next Itern' button to run simulations.