def __init__(self, ns=0, MLP=None): QObject.__init__(self) self.MLP = MLP self.interrupted = False self.layers = None if self.MLP: self.layers = self.getMlpTopology() self.ns = ns # Neighbourhood size of training rasters. self.data = None # Training data self.catlist = None # List of unique output values of the output raster self.train_error = None # Error on training set self.val_error = None # Error on validation set self.minValError = None # The minimum error that is achieved on the validation set self.valKappa = 0 # Kappa on on the validation set self.sampler = None # Sampler # Results of the MLP prediction self.prediction = None # Raster of the MLP prediction results self.confidence = None # Raster of the MLP results confidence (1 = the maximum confidence, 0 = the least confidence) self.transitionPotentials = None # Dictionary of transition potencial maps: {category1: map1, category2: map2, ...} # Outputs of the activation function for small and big numbers self.sigmax, self.sigmin = sigmoid(100), sigmoid( -100) # Max and Min of the sigmoid function self.sigrange = self.sigmax - self.sigmin # Range of the sigmoid
def __init__(self, ns=0, MLP=None): QObject.__init__(self) self.MLP = MLP self.interrupted = False self.layers = None if self.MLP: self.layers = self.getMlpTopology() self.ns = ns # Neighbourhood size of training rasters. self.data = None # Training data self.catlist = None # List of unique output values of the output raster self.train_error = None # Error on training set self.val_error = None # Error on validation set self.minValError = None # The minimum error that is achieved on the validation set self.valKappa = 0 # Kappa on on the validation set self.sampler = None # Sampler # Results of the MLP prediction self.prediction = None # Raster of the MLP prediction results self.confidence = None # Raster of the MLP results confidence (1 = the maximum confidence, 0 = the least confidence) self.transitionPotentials = None # Dictionary of transition potencial maps: {category1: map1, category2: map2, ...} # Outputs of the activation function for small and big numbers self.sigmax, self.sigmin = sigmoid(100), sigmoid(-100) # Max and Min of the sigmoid function self.sigrange = self.sigmax - self.sigmin # Range of the sigmoid
def __init__(self, ns=0, MLP=None): QObject.__init__(self) self.MLP = MLP self.layers = None if self.MLP: self.layers = self.getMlpTopology() self.ns = ns # Neighbourhood size of training rasters. self.data = None # Training data self.classlist = None # List of unique output values of the output raster self.train_error = None # Error on training set self.val_error = None # Error on validation set self.minValError = None # The minimum error that is achieved on the validation set # Results of the MLP prediction self.prediction = None # Raster of the MLP prediction results self.confidence = None # Raster of the MLP results confidence # Outputs of the activation function for small and big numbers self.sigmax, self.sigmin = sigmoid(100), sigmoid(-100) # Max and Min of the sigmoid function self.sigrange = self.sigmax - self.sigmin # Range of the sigmoid