Create app.py

app.py

@@ -0,0 +1,237 @@
+fashion_mnist = keras.datasets.fashion_mnist
+(x_train_full, y_train_full), (x_test, y_test) = fashion_mnist.load_data()
+
+x_valid, x_train = x_train_full[:5000], x_train_full[5000:]
+y_valid, y_train = y_train_full[:5000], y_train_full[5000:]
+
+x_train.shape
+import matplotlib as mpl
+import matplotlib.pyplot as plt
+plt.figure(figsize=(14,12))
+
+plt.subplot(3,3,1)
+some_image = x_train[0]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,2)
+some_image = x_train[1]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,3)
+some_image = x_train[2]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,4)
+some_image = x_train[3]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,5)
+some_image = x_train[4]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,6)
+some_image = x_train[5]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,7)
+some_image = x_train[6]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,8)
+some_image = x_train[7]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.subplot(3,3,9)
+some_image = x_train[8]
+plt.imshow(some_image, cmap=mpl.cm.binary)
+
+plt.show()
+
+class_names = ["T-shirt/top","Trouser","Pullover", "Dress","Coat","Sandals","Shirt","Sneaker","Bag","Ankle boot"]
+class_names[y_train[3]]
+
+pd_y_train = pd.DataFrame(y_train)
+
+frequency = pd_y_train.value_counts()
+category = frequency.index.tolist()
+counts = frequency.values.tolist()
+# Visualization of train set
+frequency.plot(kind='bar')
+plt.title ('Bar plot')
+plt.xlabel ('Category')
+plt.ylabel ('Frequency')
+
+img_shape = x_train.shape
+n_samples = img_shape[0]
+width = img_shape[1]
+height = img_shape[2]
+
+print("n_samples: ",n_samples)
+print("width: ",width)
+print("height: ",height)
+
+#flatten each 2d mnist image into 1d array and checing dimensions 
+x_train_flatten = x_train.reshape(n_samples, width*height)
+print("x_train_flatten.shape: ",x_train_flatten.shape)
+from sklearn.preprocessing import StandardScaler
+from sklearn.neighbors import KNeighborsClassifier
+
+# feature scaling
+standardscaler = StandardScaler() 
+X_train_scale = standardscaler.fit_transform(x_train_flatten)
+
+# import KNN classifier from sklearn 
+KNN_classifier_scale = KNeighborsClassifier(n_neighbors=5)
+KNN_classifier_scale. fit(X_train_scale,y_train)
+
+X_test_stand = standardscaler.transform(x_test_flatten)
+y_pred = KNN_classifier_scale.predict(X_test_stand)
+# Cross validation 
+from sklearn.model_selection import cross_val_score
+from sklearn.neighbors import KNeighborsClassifier
+
+# define one KNN model
+KNN_classifier = KNeighborsClassifier(n_neighbors=5, metric = 'euclidean')
+
+# call cross-val_score
+CV_scores = cross_val_score(estimator = KNN_classifier, X = x_train_flatten, y = y_train, cv = 3, scoring = 'accuracy')
+print("CV_scores: ", CV_scores)
+# Training 
+
+from sklearn.model_selection import cross_val_score
+from sklearn.neighbors import KNeighborsClassifier 
+
+import time
+start = time.time()
+
+KNN_classifier = KNeighborsClassifier(n_neighbors=5, metric = 'euclidean')
+KNN_classifier.fit(x_train_flatten, y_train)
+y_valid_predicted_label = KNN_classifier.predict(x_valid_flatten)
+
+end = time.time()
+time_duration = end-start
+print("Program finishes in {} seconds:".format(time_duration))
+# Saving the data
+from joblib import dump, load
+dump(KNN_classifier, 'KNN_fashionmnist.joblib') 
+# loading the data
+KNN_classifier = load('KNN_fashionmnist.joblib')
+
+# organize the predicted classes and actual classes into Pandas dataframe
+summary = pd.DataFrame({'predection':y_valid_predicted_label,'Original':y_valid})
+summary
+
+# Overall accuracy of the validation predictions
+from sklearn import metrics
+metrics.accuracy_score(y_valid,y_valid_predicted_label)
+
+# Calculate the per-class accuracy of the predictions
+from sklearn.metrics import confusion_matrix
+matrix = confusion_matrix(y_valid,y_valid_predicted_label)
+accuracy_score = matrix.diagonal()/matrix.sum(axis=1)
+
+print('accuracy of t-shirt is',accuracy_score[0])
+print('accuracy of Trouser is',accuracy_score[1])
+print('accuracy of pullover is',accuracy_score[2])
+print('accuracy of Dress is',accuracy_score[3])
+print('accuracy of coat is',accuracy_score[4])
+print('accuracy of sandal is',accuracy_score[5])
+print('accuracy of shirt is',accuracy_score[6])
+print('accuracy of sneaker is',accuracy_score[7])
+print('accuracy of bag is',accuracy_score[8])
+print('accuracy of boot is',accuracy_score[9])
+
+# visualize the classification confusion matrix to check the details of the validation predictions for each class
+import matplotlib.pyplot as plt
+from sklearn.metrics import ConfusionMatrixDisplay
+ConfusionMatrixDisplay.from_predictions(y_valid, y_valid_predicted_label)
+plt.title("Classification Confusion matrix")
+plt.show()
+
+# Task 4.1.9 Different K values, and select the best model that has highest validation accuracy
+# visualize the classification confusion matrix on the test set to report the details of predictions over every class
+from sklearn.neighbors import KNeighborsClassifier
+KNN_classifier = KNeighborsClassifier(n_neighbors=3, metric = 'euclidean')
+KNN_classifier.fit(x_train_flatten, y_train)
+y_valid_predicted_label = KNN_classifier.predict(x_valid_flatten)
+
+y_valid_predicted_label
+from sklearn import metrics
+metrics.accuracy_score(y_valid, y_valid_predicted_label)
+
+import matplotlib.pyplot as plt
+from sklearn.metrics import ConfusionMatrixDisplay
+ConfusionMatrixDisplay.from_predictions(y_test, y_test_pred)
+plt.title("classifiaction confusion matrix")
+plt.show()
+# Task 4.1.10: Calculate the overall accuracy of the predictions over  validation set and test set using the best model 
+from sklearn import metrics
+metrics.accuracy_score(y_test, y_test_pred)
+# discriminant analysis
+import numpy as np
+from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
+clf = LinearDiscriminantAnalysis()
+clf.fit(x_train_flatten, y_train)
+
+start = time.time()
+predicted_labels = clf.predict(x_valid_flatten)
+end = time.time()
+time_duration = end-start
+print("Program finishes in {} seconds:".format(time_duration))
+
+y_test_pred = clf.predict(x_test_flatten)
+print("Accuracy of testing Set: ", metrics.accuracy_score(y_test, y_test_pred))
+y_valid_pred = clf.predict(x_valid_flatten)
+print("Accuracy of validation Set: ", metrics.accuracy_score(y_valid, y_valid_pred))
+
+from sklearn.metrics import confusion_matrix
+matrix = confusion_matrix(y_test,y_test_pred)
+accuracy_score = matrix.diagonal()/matrix.sum(axis=1)
+
+print('accuracy of t-shirt is',accuracy_score[0])
+print('accuracy of Trouser is',accuracy_score[1])
+print('accuracy of pullover is',accuracy_score[2])
+print('accuracy of Dress is',accuracy_score[3])
+print('accuracy of coat is',accuracy_score[4])
+print('accuracy of sandal is',accuracy_score[5])
+print('accuracy of shirt is',accuracy_score[6])
+print('accuracy of sneaker is',accuracy_score[7])
+print('accuracy of bag is',accuracy_score[8])
+print('accuracy of boot is',accuracy_score[9])
+
+from gradio.outputs import Label
+import gradio as gr
+import cv2
+import tensorflow as tf
+
+def caption(image,input_module1):
+  class_names = ["T-shirt/top", "Trouser", "Pullover", "Dress", "Coat",
+  "Sandal", "Shirt", "Sneaker", "Bag", "Ankle boot"] 
+  image=image.reshape(1,28*28)
+  if input_module1=="KNN":
+    output1=KNN_classifier.predict(image)[0]
+    predictions=KNN_classifier.predict_proba(image)[0]
+    
+  elif input_module1==("Linear discriminant analysis"):
+    output1=clf.predict(image)[0]
+    predictions=clf.predict_proba(image)[0]
+    
+  elif input_module1==("Quadratic discriminant analysis"):
+    output1=qda.predict(image)[0]
+    predictions=qda.predict_proba(image)[0]
+    
+  elif input_module1=="Naive Bayes classifier":
+    output1=gnb.predict(image)[0]
+    predictions=gnb.predict_proba(image)[0]
+  
+  output2 = {}
+
+  for i in range(len(predictions)):
+    output2[class_names[i]] = predictions[i]
+  return output1 ,output2
+
+input_module = gr.inputs.Image(label = "Input Image",image_mode="L",shape=(28,28))
+input_module1 = gr.inputs.Dropdown(choices=["KNN","Linear discriminant analysis", "Quadratic discriminant analysis","Naive Bayes classifier"], label = "Method")
+output1 = gr.outputs.Textbox(label = "Predicted Class")
+output2=gr.outputs.Label(label= "probability of class")
+gr.Interface(fn=caption, inputs=[input_module,input_module1], outputs=[output1,output2]).launch(debug=True)