snatch59
diff --git a/‎README.md‎
Lines changed: 22 additions & 4 deletions b/‎README.md‎
Lines changed: 22 additions & 4 deletions
diff --git a/‎assets/gnb_cm.png‎
116 KB b/‎assets/gnb_cm.png‎
116 KB
diff --git a/‎assets/lda_cm.png‎
116 KB b/‎assets/lda_cm.png‎
116 KB
diff --git a/‎assets/qda_cm.png‎
120 KB b/‎assets/qda_cm.png‎
120 KB
diff --git a/‎inceptionv3_svm_classifier.py‎
Lines changed: 30 additions & 0 deletions b/‎inceptionv3_svm_classifier.py‎
Lines changed: 30 additions & 0 deletions
@@ -23,12 +23,15 @@ following:
 * Random Forest (RF)
 * K-Nearest Neighbor (KNN)
 * Multi-Layer Perceptron (ML)
+* Gaussian Naive Bayes (GNB)
+* Linear Discriminant Analysis (LDA)
+* Quadratic Discriminant Analysis (QDA)
 
 Training and validation time, and the accuracy of each classifier is
-displayed. Most classifier were run with their default tuning values,
-however tuning was carried on those classifier that fell well below 90%
-accuracy for their defaults, such of Extra Trees and Random Forsest
-(initially in the 75 - 78% region).
+displayed. Most classifiers were run with their default tuning values,
+however tuning was carried, where possible, on those classifiers that
+fell well below 90% accuracy for their defaults, such of Extra Trees
+and Random Forsest (initially in the 75 - 78% region).
 
 A summary of the results is as follows (training/test time, accuracy):
 
@@ -37,6 +40,9 @@ A summary of the results is as follows (training/test time, accuracy):
 * RF: 16.47 sec, 90.8%
 * KNN: 2.2 sec, 91.5%
 * MLP: 13.83 sec, 97.1%
+* GNB: 1.1 sec, 91.8%
+* LDA: 4.95 sec, 91.0%
+* QDA: 0.84 sec, 5.3% (Variables are collinear warning!)
 
 Note that these results vary between runs, and are just representative.
 
@@ -80,3 +86,15 @@ laptop.
 ## Multi-Layer Perceptron
 
 ![caltech MLP confusion matrix](assets/mlp_cm.png)
+
+## Gaussian Naive Bayes
+
+![caltech GNB confusion matrix](assets/gnb_cm.png)
+
+## Linear Discriminant Analysis
+
+![caltech LDA confusion matrix](assets/lda_cm.png)
+
+## Quadratic Discriminant Analysis
+
+![caltech QDA confusion matrix](assets/qda_cm.png)
@@ -7,6 +7,9 @@
 from sklearn.ensemble import RandomForestClassifier, ExtraTreesClassifier
 from sklearn.neighbors import KNeighborsClassifier
 from sklearn.neural_network import MLPClassifier
+from sklearn.naive_bayes import GaussianNB
+from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
+from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis
 import numpy as np
 import matplotlib.pyplot as plt
 import pickle
@@ -231,3 +234,30 @@ def run_classifier(clfr, x_train_data, y_train_data, x_test_data, y_test_data, a
 clf = MLPClassifier()
 run_classifier(clf, X_train, y_train, X_test, y_test, "CNN-MLP Accuracy: {0:0.1f}%",
                "Multi-layer Perceptron Confusion matrix")
+
+# GaussianNB defaults:
+# priors=None
+
+# classify the images with a Gaussian Naive Bayes Classifier
+print('Gaussian Naive Bayes Classifier starting ...')
+clf = GaussianNB()
+run_classifier(clf, X_train, y_train, X_test, y_test, "CNN-GNB Accuracy: {0:0.1f}%",
+               "Gaussian Naive Bayes Confusion matrix")
+
+# LinearDiscriminantAnalysis defaults:
+# solver=’svd’, shrinkage=None, priors=None, n_components=None, store_covariance=False, tol=0.0001
+
+# classify the images with a Quadratic Discriminant Analysis Classifier
+print('Linear Discriminant Analysis Classifier starting ...')
+clf = LinearDiscriminantAnalysis()
+run_classifier(clf, X_train, y_train, X_test, y_test, "CNN-LDA Accuracy: {0:0.1f}%",
+               "Linear Discriminant Analysis Confusion matrix")
+
+# QuadraticDiscriminantAnalysis defaults:
+# priors=None, reg_param=0.0, store_covariance=False, tol=0.0001, store_covariances=None
+
+# classify the images with a Quadratic Discriminant Analysis Classifier
+print('Quadratic Discriminant Analysis Classifier starting ...')
+clf = QuadraticDiscriminantAnalysis()
+run_classifier(clf, X_train, y_train, X_test, y_test, "CNN-QDA Accuracy: {0:0.1f}%",
+               "Quadratic Discriminant Analysis Confusion matrix")