Perceptron

An implementation of a perceptron -a single level neural network- for classification.

:- use_module('path/to/.../src/methods/perceptron/perceptron.pl').

%% usage example
TrainData = [5.1,3.5,1.4, 4.9,3.0,1.4, 4.7,3.2,1.3, 4.6,3.1,1.5],
TestData = [3,2,0, 5,1,4, 0,0,4, 3,3,5, 0,5,5, 2,5,5],
perceptron_initModelWithTrain(TrainData, 3, [0,1,0,1], 2, 1000),
perceptron_classify(TestData, 3, PredictList).

Available Predicates

links/resources

perceptron_initModelNoTrain/3

Initilizes the perceptron model and its weight matrix but doesn't train it.

%% predicate definition 
perceptron_initModelNoTrain(NumClasses, Dimensionality, MaxIterations) :-
        NumClasses >= 0,
        Dimensionality > 0,
        MaxIterations >= 0,
        initModelNoTrainI(NumClasses, Dimensionality, MaxIterations).

%% foreign c++ predicate definition 
foreign(initModelNoTrain, c, initModelNoTrainI(+integer, +integer, +integer)).

Parameters

Name	Type	Description	Default
numClasses	+integer		0
dimensionality	+integer	Number of classes in the data.	0
maxIterations	+integer	The maximum number of iterations the perceptron is to be run.	1000

perceptron_initModelWithTrain/5

Initilizes the perceptron model and its weight matrix and trains it with the given data.

%% predicate definition 
perceptron_initModelWithTrain(DataList, DataRows, LabelsList, NumClasses, MaxIterations) :-
        NumClasses >= 0,
        MaxIterations >= 0,
        convert_list_to_float_array(DataList, DataRows, array(Xsize, Xrownum, X)),
        convert_list_to_float_array(LabelsList, array(Ysize, Y)),
        initModelWithTrainI(X, Xsize, Xrownum, Y, Ysize, NumClasses, MaxIterations).

%% foreign c++ predicate definition 
foreign(initModelWithTrain, c, initModelWithTrainI( +pointer(float_array), +integer, +integer, 
                                                    +pointer(float_array), +integer, 
                                                    +integer, +integer)).

Parameters

Name	Type	Description	Default
data	+matrix	A matrix containing the training set.	-
labels	+vector	A vector containing labels for the training set.	-
numClasses	+integer	Number of classes in the data.	0
maxIterations	+integer	The maximum number of iterations the perceptron is to be run.	1000

perceptron_biases/1

Get the biases.

%% predicate definition 
perceptron_biases(BiasesList) :-
        biasesI(Y, Ysize),
        convert_float_array_to_list(Y, Ysize, BiasesList).

%% foreign c++ predicate definition 
foreign(biases, c, biasesI(-pointer(float_array), -integer)).

Parameters

Name	Type	Description	Default
biases	-vector	Get the biases.	-

perceptron_classify/3

After training, use the weights matrix to classify test, and put the predicted classes in predictedLabels.

%% predicate definition 
perceptron_classify(TestList, TestRows, PredictLabelList) :-
        convert_list_to_float_array(TestList, TestRows, array(Xsize, Xrows, X)),
        classifyI(X, Xsize, Xrows, Y, Ysize),
        convert_float_array_to_list(Y, Ysize, PredictLabelList).

%% foreign c++ predicate definition 
foreign(classify, c, classifyI( +pointer(float_array), +integer, +integer, 
                                -pointer(float_array), -integer)).

Parameters

Name	Type	Description	Default
test	+matrix		-
predicLabels	-vector	The predicted labels for the test data.	-

perceptron_train/5

Train the perceptron on the given data for up to the maximum number of iterations. This training does not reset the model weights, so you can call train/8 on multiple datasets sequentially.

%% predicate definition 
perceptron_train(DataList, DataRows, LabelsList, NumClasses, WeightsList) :-
        NumClasses >= 0,
        convert_list_to_float_array(DataList, DataRows, array(Xsize, Xrownum, X)),
        convert_list_to_float_array(LabelsList, array(Ysize, Y)),
        convert_list_to_float_array(WeightsList, array(Zsize, Z)),
        trainI(X, Xsize, Xrownum, Y, Ysize, NumClasses, Z, Zsize).

%% foreign c++ predicate definition 
foreign(train, c, trainI( +pointer(float_array), +integer, +integer, 
                          +pointer(float_array), +integer, 
                          +integer, 
                          +pointer(float_array), +integer)).

Parameters

Name	Type	Description	Default
data	+matrix	A matrix containing the training set.	-
labels	+vector	A vector containing labels for the training set.	-
numClasses	+integer	Number of classes in the data.	0
instanceWeights	+vector	Cost matrix. Stores the cost of mispredicting instances. This is useful for boosting.	-

perceptron_weights/2

Get the weight matrix.

%% predicate definition 
perceptron_weights(WeightsList, XCols) :-
        weightsI(X, XCols, XRows),
        convert_float_array_to_2d_list(X, XCols, XRows, WeightsList).

%% foreign c++ predicate definition 
foreign(weights, c, weightsI(-pointer(float_array), -integer, -integer)).

Parameters

Name	Type	Description	Default
weights	-matrix	Get the weight matrix.	-

Connected Links/Resources

If you want a more detailed explanation, then go to the python documentation. There is most of the time a good explanation on how the methods work and what the parameters do.

added some of the links from the python documentation

Comments

Please register or sign in to add a comment.