void SearchSolution()
{
// create network
DistanceNetwork network = new DistanceNetwork(2, neurons);
// set random generators range
foreach (var neuron in network.Layers.SelectMany(layer => layer?.Neurons).Where(neuron => neuron != null))
{
neuron.RandGenerator = new UniformContinuousDistribution(new Range(0, 1000));
}
// create learning algorithm
ElasticNetworkLearning trainer = new ElasticNetworkLearning(network);
double fixedLearningRate = learningRate / 20;
double driftingLearningRate = fixedLearningRate * 19;
double[,] path = new double[neurons + 1, 2];
double[] input = new double[2];
int i = 0;
while (!needToStop)
{
// update learning speed & radius
trainer.LearningRate = driftingLearningRate * (iterations - i) / iterations + fixedLearningRate;
trainer.LearningRadius = learningRadius * (iterations - i) / iterations;
// set network input
int currentCity = rand.Next(citiesCount);
input[0] = map[currentCity, 0];
input[1] = map[currentCity, 1];
// run one training iteration
trainer.Run(input);
// show current path
for (int j = 0; j < neurons; j++)
{
path[j, 0] = network.Layers[0].Neurons[j].Weights[0];
path[j, 1] = network.Layers[0].Neurons[j].Weights[1];
}
path[neurons, 0] = network.Layers[0].Neurons[0].Weights[0];
path[neurons, 1] = network.Layers[0].Neurons[0].Weights[1];
chart.UpdateDataSeries("path", path);
i++;
SetText(currentIterationBox, i.ToString());
if (i >= iterations)
{
break;
}
}
// enable settings controls
EnableControls(true);
}
// Worker thread
void SearchSolution()
{
// set random generators range
Neuron.RandRange = new Range(0, 1000);
// create network
var network = new DistanceNetwork(2, neurons);
// create learning algorithm
var trainer = new ElasticNetworkLearning(network);
var fixedLearningRate = learningRate / 20;
var driftingLearningRate = fixedLearningRate * 19;
// path
var path = new double[neurons + 1, 2];
// input
var input = new double[2];
// iterations
var i = 0;
// loop
while (!needToStop)
{
// update learning speed & radius
trainer.LearningRate = driftingLearningRate * (iterations - i) / iterations + fixedLearningRate;
trainer.LearningRadius = learningRadius * (iterations - i) / iterations;
// set network input
var currentCity = rand.Next(citiesCount);
input[0] = map[currentCity, 0];
input[1] = map[currentCity, 1];
// run one training iteration
trainer.Run(input);
// show current path
for (var j = 0; j < neurons; j++)
{
path[j, 0] = network.Layers[0].Neurons[j].Weights[0];
path[j, 1] = network.Layers[0].Neurons[j].Weights[1];
}
path[neurons, 0] = network.Layers[0].Neurons[0].Weights[0];
path[neurons, 1] = network.Layers[0].Neurons[0].Weights[1];
chart.UpdateDataSeries("path", path);
// increase current iteration
i++;
// set current iteration's info
SetText(currentIterationBox, i.ToString());
// stop ?
if (i >= iterations)
{
break;
}
}
// enable settings controls
EnableControls(true);
}
//http://www.aforgenet.com/framework/docs/
void SearchSolution()
{
// set random generators range
Neuron.RandRange = new DoubleRange(0, 1000);
// create network
DistanceNetwork network = new DistanceNetwork(2, neurons);
// create learning algorithm
ElasticNetworkLearning trainer = new ElasticNetworkLearning(network);
double fixedLearningRate = learningRate / 20;
double driftingLearningRate = fixedLearningRate * 19;
// path
double[,] path = new double[neurons + 1, 2];
// input
double[] input = new double[2];
// iterations
int i = 0;
// loop
while (!needToStop)
{
// update learning speed & radius
trainer.LearningRate = driftingLearningRate * (iterations - i) / iterations + fixedLearningRate;
trainer.LearningRadius = learningRadius * (iterations - i) / iterations;
// set network input
int currentCity = rand.Next(citiesCount);
input[0] = map[currentCity, 0];
input[1] = map[currentCity, 1];
// run one training iteration
trainer.Run(input);
// show current path
for (int j = 0; j < neurons; j++)
{
path[j, 0] = network[0][j][0];
path[j, 1] = network[0][j][1];
}
path[neurons, 0] = network[0][0][0];
path[neurons, 1] = network[0][0][1];
chart1.UpdateDataSeries("Caminho", path);
i++;
SetText(i.ToString());
// stop ?
if (i >= iterations)
{
break;
}
}
}
