/// <summary>
/// Runs through the network and makes a prediction.
/// </summary>
/// <param name="input">The input image to run the prediction on.</param>
/// <returns>The values from the final layer. The largest value is the networks prediction.</returns>
public double[] FeedForward(double[] input)
{
Matrix <double> _currentImages = CreateMatrix.DenseOfRowVectors(CreateVector.DenseOfArray <double>(input));
foreach (ILayerInformation _layerInformation in this.LayerInformation)
{
switch (_layerInformation.LayerType)
{
case (LayerType.Convolutional):
ConvolutionalLayerInformation _convInfo = _layerInformation as ConvolutionalLayerInformation;
_currentImages = this.Convolve(_convInfo, _currentImages);
break;
case (LayerType.Pooling):
PoolingLayerInformation _poolInfo = _layerInformation as PoolingLayerInformation;
_currentImages = this.Pool(_poolInfo, _currentImages);
break;
case (LayerType.NonLinear):
NonLinearLayerInformation _nonLinearInfo = _layerInformation as NonLinearLayerInformation;
_currentImages = this.NonLinear(_nonLinearInfo, _currentImages);
break;
}
}
double[] _fullyConnectedInput = CreateVector.DenseOfEnumerable <double>(_currentImages.EnumerateRows().SelectMany(a => a)).ToArray();
return(this.FullyConnectedNetwork.FeedForward(_currentImages.Enumerate().ToArray()));
}
/// <summary>
/// Runs through a single pooling layer.
/// </summary>
/// <param name="layerInfo">The layer info used for this layer.</param>
/// <param name="inputImages">
/// A matrix of all the images that will be pooled. Each row is an image.
/// </param>
/// <returns>A matrix of all the resulting images. Each row is an image.</returns>
internal Matrix <double> Pool(PoolingLayerInformation layerInfo, Matrix <double> inputImages)
{
Matrix <double> _preConvolutionMap = null;
Matrix <double> _outputImages = CreateMatrix.Dense <double>(inputImages.RowCount, (int)Math.Pow((Math.Sqrt(inputImages.ColumnCount) - layerInfo.KernelSize) / layerInfo.Stride + 1, 2));
for (int i = 0; i < inputImages.RowCount; i++)
{
_preConvolutionMap = this.CreateMaskingMap(layerInfo.KernelSize, layerInfo.Stride, inputImages.Row(i));
switch (layerInfo.PoolingType)
{
case (PoolingType.MaxPooling):
_outputImages.SetRow(i, _preConvolutionMap.ColumnNorms(double.PositiveInfinity));
break;
}
}
return(_outputImages);
}
