8382_Ershov_Mikhail_lb7

8382/Ershov/lb/7/main.py

@@ -0,0 +1,99 @@
+import numpy as np
+from keras.datasets import imdb
+from keras.models import Sequential
+from keras.layers import Dense
+from keras.layers import LSTM,Conv1D,MaxPooling1D,Dropout
+from keras.layers.embeddings import Embedding
+from keras.preprocessing import sequence
+import matplotlib.pyplot as plt
+
+top_words = 10000
+
+(training_data, training_targets), (testing_data, testing_targets) = imdb.load_data(num_words=10000)
+data = np.concatenate((training_data, testing_data), axis=0)
+targets = np.concatenate((training_targets, testing_targets), axis=0)
+
+max_review_length = 500
+training_data = sequence.pad_sequences(training_data, maxlen=max_review_length)
+testing_data = sequence.pad_sequences(testing_data, maxlen=max_review_length)
+embedding_vecor_length = 32
+
+
+def build_model_1():
+    model = Sequential()
+    model.add(Embedding(top_words, embedding_vecor_length, input_length=max_review_length))
+    model.add(Dropout(0.2))
+    model.add(LSTM(100))
+    model.add(Dropout(0.2))
+    model.add(Dense(1, activation='sigmoid'))
+    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
+    return model
+
+
+def build_model_2():
+    model = Sequential()
+    model.add(Embedding(top_words, embedding_vecor_length, input_length=max_review_length))
+    model.add(Conv1D(filters=32, kernel_size=3, padding='same', activation='relu'))
+    model.add(MaxPooling1D(pool_size=2))
+    model.add(Dropout(0.2))
+    model.add(LSTM(100))
+    model.add(Dropout(0.2))
+    model.add(Dense(1, activation='sigmoid'))
+    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
+    return model
+
+
+def vectorize(sequences, dimension=500):
+    results = np.zeros((len(sequences), dimension))
+    for i, sequence in enumerate(sequences):
+        results[i, sequence] = 1
+    return results
+
+
+def test_text(filename):
+    f = open(filename, 'r')
+    open_text = f.read()
+    # print(open_text)
+    index = imdb.get_word_index()
+    text = []
+    for i in open_text:
+        if i in index and index[i] < 500:
+            text.append(index[i])
+
+    text = vectorize([text])
+    return text
+
+
+def final_pred(model_1,model_2,text):
+    pred_1 = model_1.predict(text)
+    pred_2 = model_2.predict(text)
+    result = 0.5*(pred_1+pred_2)
+    if result >= 0.5:
+        print('Отзыв хороший')
+    else:
+        print('Отзыв плохой')
+
+
+def evaluation(model_1,model_2,testing_data):
+    scores_1 = model_1.predict(testing_data)
+    scores_2 = model_2.predict(testing_data)
+    res = 0.5*(scores_1+scores_2)
+    res = round(res)
+    return res
+
+
+if __name__ == '__main__':
+    model_1 = build_model_1()
+    model_1.fit(training_data, training_targets, validation_data=(testing_data, testing_targets), epochs=3, batch_size=64)
+    scores_1 = model_1.evaluate(testing_data, testing_targets, verbose=0)
+    model_2 = build_model_2()
+    model_2.fit(training_data, training_targets, validation_data=(testing_data, testing_targets), epochs=3, batch_size=64)
+    scores_2 = model_2.evaluate(testing_data, testing_targets, verbose=0)
+    print("Accuracy: %.2f%%" % (scores_1[1] * 100))
+    print("Accuracy: %.2f%%" % (scores_2[1] * 100))
+    result = evaluation(model_1,model_2,testing_data)
+
+    filename = "text_good.txt"
+    print(filename)
+    text = test_text(filename)
+    final_pred(model_1,model_2,text)

8382/Ershov/lb/7/text_good.txt

@@ -0,0 +1 @@
+A brilliant masterpiece with unique style. It starts rather slow, but half time passed it burns the ground. It has everithing you want. Humour, Fantastic style, twisted plot, cool and memorable characters and action.