Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
5 changed files
with
246 additions
and
79 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Binary file not shown.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,171 @@ | ||
| __version__= '0.1' | ||
|
|
||
| __author__ = 'Sagar Gautam' | ||
| import math | ||
|
|
||
| import jpype | ||
|
|
||
| from jpype import * | ||
|
|
||
| from pybrain.tools.shortcuts import buildNetwork | ||
|
|
||
| from pybrain.tools.xml.networkreader import NetworkReader | ||
|
|
||
| from pybrain.structure import TanhLayer, SigmoidLayer | ||
|
|
||
| # Normalize result | ||
| def normalize(output): | ||
| if output[0] > output[1]: | ||
| return [1, 0] | ||
| else: | ||
| return [0, 1] | ||
|
|
||
| # Measure euclidean distance between feature vectors | ||
| def edit_distance(feature1, feature2): | ||
| total=0 | ||
| for i in range(len(feature1)): | ||
| total = total+(feature1[i]-feature2[i])*(feature1[i]-feature2[i]) | ||
| distance = math.sqrt(total) | ||
| return distance | ||
|
|
||
| # Assign majority class cabel to the label of new instance | ||
| def get_label(labels): | ||
| count1 = 0 | ||
| for label in range(len(labels)): | ||
| if labels[label] == 1: | ||
| count1 = count1 + 1 | ||
| count0 = len(labels)-count1 | ||
|
|
||
| if count1 > count0: | ||
| new_label = 1 | ||
| else: | ||
| new_label = 0 | ||
| return new_label | ||
|
|
||
| class Model: | ||
| test_features = [] | ||
| train = [] | ||
| demented_file_no = 75 | ||
| non_demented_file_no = 107 | ||
|
|
||
| def __init__(self,Source): | ||
| self.Source = Source | ||
| if isJVMstarted(): | ||
|
|
||
| startJVM(jpype.getDefaultJVMPath()); | ||
| myPackage = JPackage('FeatureExtraction').org.classes | ||
| dementia = myPackage.Dementia | ||
| string = dementia.System("Test",Source) | ||
| test_features = [float(x) for x in string.split()] | ||
| shutdownJVM(); | ||
|
|
||
| def KNN(self,K): | ||
| for i in range(0,demented_file_no): | ||
| filename = 'Resource/TrainData/Demented/Nearest/Image'+str(i)+'.txt' | ||
| data = [float(x) for x in open(filename).read().split()] | ||
| data_sample = (tuple(data), 0) | ||
| train.append(data_sample) | ||
|
|
||
| for j in range(0,non_demented_file_no): | ||
| filename = 'Resource/TrainData/NotDemented/Nearest/Image'+str(j)+'.txt' | ||
| data = [float(x) for x in open(filename).read().split()] | ||
| data_sample = (tuple(data), 1) | ||
| train.append(data_sample) | ||
|
|
||
| test_feature = tuple(test_features) | ||
| print test_feature | ||
| distance_vector = [1.0e+15]*K | ||
| label_vector = [1]*K | ||
|
|
||
| # KNN alorithm to test the new instances | ||
|
|
||
| for feature, label in train: | ||
| distance = edit_distance(feature, test_feature) | ||
| for value in range(K): | ||
| if distance < distance_vector[value]: | ||
| distance_vector[value] = distance | ||
| label_vector[value] = label | ||
| break | ||
|
|
||
| # Assign majority label to label of test feature vector | ||
| new_label = get_label(label_vector) | ||
| if new_label == 0: | ||
| print "KNN Model: Dementia Positive" | ||
| if new_label == 1: | ||
| print "KNN Model: Dementia Negative" | ||
|
|
||
| #def ANN(self): | ||
|
|
||
| # Acess java class Dementia from package featureextraction to calculate feature fector | ||
|
|
||
| startJVM(jpype.getDefaultJVMPath()); | ||
| myPackage = JPackage('FeatureExtraction').org.classes # Location of Dementia.class | ||
| dementia = myPackage.Dementia # Dementia.class file acess | ||
| source = raw_input("Enter Test Image File Location:") | ||
| string = dementia.System("Test",source) # Static function System(....,....) call | ||
| features = [float(x) for x in string.split()] # We get feature vector as string so to get numeric | ||
| # Spliting is required | ||
|
|
||
| # Best Approximated Value of Nearest Neighbours | ||
| K = 7 | ||
|
|
||
| demented_file_no = 75 | ||
| non_demented_file_no = 107 | ||
|
|
||
| # List of Training Data of type [(_,_,_),_] | ||
| train = [] | ||
| # List of Test Data of type [(_,_,_),_] | ||
|
|
||
| for i in range(0,demented_file_no): | ||
| filename = 'Resource/TrainData/Demented/Nearest/Image'+str(i)+'.txt' | ||
| data = [float(x) for x in open(filename).read().split()] | ||
| data_sample = (tuple(data), 0) | ||
| train.append(data_sample) | ||
|
|
||
| for j in range(0,non_demented_file_no): | ||
| filename = 'Resource/TrainData/NotDemented/Nearest/Image'+str(j)+'.txt' | ||
| data = [float(x) for x in open(filename).read().split()] | ||
| data_sample = (tuple(data), 1) | ||
| train.append(data_sample) | ||
|
|
||
| test_feature = tuple(features) | ||
| distance_vector = [1.0e+15]*K | ||
| label_vector = [1]*K | ||
|
|
||
| # KNN alorithm to test the new instances | ||
|
|
||
| for feature, label in train: | ||
| distance = edit_distance(feature, test_feature) | ||
| for value in range(K): | ||
| if distance < distance_vector[value]: | ||
| distance_vector[value] = distance | ||
| label_vector[value] = label | ||
| break | ||
|
|
||
| # Assign majority label to label of test feature vector | ||
| new_label = get_label(label_vector) | ||
|
|
||
| if new_label == 0: | ||
| print "KNN Model: Dementia Positive" | ||
|
|
||
| if new_label == 1: | ||
| print "KNN Model: Dementia Negative" | ||
| shutdownJVM(); | ||
| model = Model("Resource/Sample/Image5.img") | ||
| model.KNN() | ||
| # Network initialization | ||
| network = buildNetwork(3, 2, 2, 2, bias=True, hiddenclass=TanhLayer) | ||
|
|
||
| # Assign Weights to the neural network from xml file | ||
| network = NetworkReader.readFrom('filename.xml') | ||
|
|
||
| # Get output of the neural netwok | ||
| output = network.activate(features) | ||
| # Get final result from output | ||
| prediction = normalize(output) | ||
|
|
||
| if(prediction == [0, 1]): | ||
| print "Neural Network: Dementia Positive" | ||
| if(prediction == [1, 0]): | ||
| print "Neural Network: Dementia Negative" | ||
|
|