학교 cost function그래프 과제 질문

Question

고등학교 미분 발표 과제 cost funtion그래프가 이차함수 모양을 이룬다는 걸 보여주고, 미분을 통해 가장 적절한 값을 낼 수 있다라는 주제로 발표 하려 하는데요, from tensorflow.keras import layers, modelsfrom tensorflow.keras.datasets import cifar10from tensorflow.keras.utils import to_categoricalfrom tensorflow.keras.preprocessing import imageimport numpy as npfrom sklearn.metrics import classification_report, confusion_matriximport matplotlib.pyplot as pltimport seaborn as sns (x_train, y_train), (x_test, y_test) &#61; cifar10.load_data() class_indices &#61; [3,5]train_indices &#61; np.where(np.isin(y_train, class_indices))[0]test_indices &#61; np.where(np.isin(y_test, class_indices))[0]x_train &#61; x_train[train_indices]y_train &#61; y_train[train_indices]x_test &#61; x_test[test_indices]y_test &#61; y_test[test_indices] x_train &#61; x_train.astype('float32')/255.0x_test &#61; x_test.astype('float32')/255.0y_train &#61; (y_train &#61;&#61; 5).astype('float32')y_test &#61; (y_test &#61;&#61; 5).astype('float32')model &#61; models.Sequential()model.add(layers.Conv2D(32, (3, 3), activation &#61; 'relu', input_shape &#61; (32, 32, 3)))model.add(layers.MaxPooling2D((2, 2)))model.add(layers.Conv2D(64, (3, 3), activation &#61; 'relu'))model.add(layers.MaxPooling2D((2, 2)))model.add(layers.Conv2D(64, (3, 3), activation &#61; 'relu'))model.add(layers.Flatten())model.add(layers.Dense(64, activation &#61; 'relu'))model.add(layers.Dense(1, activation &#61; 'sigmoid'))model.compile(optimizer &#61; 'adam', loss &#61; 'binary_crossentropy', metrics &#61; ['accuracy'])history &#61; model.fit(x_train, y_train, epochs &#61; 10, batch_size &#61; 64, validation_data &#61; (x_test, y_test))def classify_image(image_path): img &#61; image.load_img(image_path, target_size &#61; (32, 32)) img_array &#61; image.img_to_array(img) img_array &#61; img_array / 255.0 img_array &#61; np.expand_dims(img_array, axis &#61; 0) prediction &#61; model.predict(img_array) predicted_class &#61; np.round(prediction[0][0]).astype(int) predicted_label &#61; '개' if predicted_class &#61;&#61; 1 else '고양이' return predicted_labelimage_path &#61; input('이미지 경로를 입력하세요 : ')predicted_label &#61; classify_image(image_path)img &#61; image.load_img(image_path)plt.imshow(img)plt.axis('off')plt.show()print(f&#34;입력된 이미지는 {predicted_label}입니다.&#34;) loss, accuracy &#61; model.evaluate(x_test, y_test)print(loss)print(accuracy) plt.plot(history.history['accuracy'], label &#61; 'Training Accuracy')plt.plot(history.history['val_accuracy'], label &#61; 'Validation Accuracy')plt.xlabel('Epoch')plt.ylabel('Accuracy')plt.legend()plt.show() predicted_prob &#61; model.predict(x_test)predicted_labels &#61; (predicted_prob > 0.5).astype('int32')print(classification_report(y_test, predicted_labels)) cm &#61; confusion_matrix(y_test, predicted_labels)plt.figure(figsize &#61; (8, 6))sns.heatmap(cm, annot &#61; True, fmt &#61; 'd', cmap &#61; 'Greys')plt.xlabel('Predicted label')plt.ylabel('True label')plt.title('Confusion Matrix Heatmap')plt.show() 간단한 개, 고양이 분류 프로그램인데, 시도 횟수에 따른 loss값을 나타낼 수는 있는데, weight에 따른 loss값을 어떻게 나타내야 할지 모르겠어서 질문 드립니다. Gpt한테도 물어봤는대 안돼가지고 여기 질문드립니다.

학교 cost function그래프 과제 질문

0개의 답변

Q&A 목록