diff --git a/Assignment 3/200816_A3.ipynb b/Assignment 3/200816_A3.ipynb new file mode 100644 index 0000000..d977d41 --- /dev/null +++ b/Assignment 3/200816_A3.ipynb @@ -0,0 +1 @@ +{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Copy of CV_with_TF-Assignment-3.ipynb","provenance":[{"file_id":"1asGptVW1kSOD-sm44G3PJr-Ad9lPwQFR","timestamp":1657696905259},{"file_id":"1D6ydyPPCMxckA2UNfQAfoDNQMednmPXO","timestamp":1655269145670},{"file_id":"1SOXxLRZADUvtLiTygZYTHBTZP2l0W_jT","timestamp":1655267663630},{"file_id":"1nTfv0Eyv59_cvAq13soX2DgTp0t4G6M3","timestamp":1655196024245}]},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU","gpuClass":"standard"},"cells":[{"cell_type":"markdown","source":["# Assignment 3\n","\n"," In this Assignment, we will use CNN to classify digits. \n","The `MNIST` database is a large database of handwritten digits that is commonly used for training various image processing systems.\n","\n"],"metadata":{"id":"VGHh_5UYzKpV"}},{"cell_type":"markdown","source":["## Importing TensorFlow"],"metadata":{"id":"JnsMbCPNzPAr"}},{"cell_type":"code","execution_count":21,"metadata":{"id":"HRLTw3cMwvi7","executionInfo":{"status":"ok","timestamp":1657696485418,"user_tz":-330,"elapsed":7,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}}},"outputs":[],"source":["import tensorflow as tf"]},{"cell_type":"markdown","source":["## Get the dataset"],"metadata":{"id":"6Ji7HGpgzSPi"}},{"cell_type":"code","source":["# Import the dataset\n","\n","(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()"],"metadata":{"id":"oEW3KDEvzIHL","executionInfo":{"status":"ok","timestamp":1657696486164,"user_tz":-330,"elapsed":4,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}}},"execution_count":22,"outputs":[]},{"cell_type":"code","source":["# Split the dataset\n","\n","from sklearn.model_selection import train_test_split\n","X_val,X_test,Y_val,Y_test=train_test_split(x_test,y_test,test_size=0.2)"],"metadata":{"id":"F_sRU9dx_mYQ","executionInfo":{"status":"ok","timestamp":1657696546445,"user_tz":-330,"elapsed":492,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}}},"execution_count":30,"outputs":[]},{"cell_type":"markdown","source":["## Visualize the dataset\n","Print some images with labels."],"metadata":{"id":"EVpQheoVqoEG"}},{"cell_type":"code","source":["# Pre processing \n","import numpy as np\n","\n","def process(dataset):\n"," return np.expand_dims((dataset.astype('float32')/255.),axis=3)\n","\n","x_train_norm = process(x_train)\n","X_val_norm = process(X_val)\n","X_test_norm = process(X_test)"],"metadata":{"id":"yF1Nj63Bz9m7","executionInfo":{"status":"ok","timestamp":1657696549452,"user_tz":-330,"elapsed":576,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}}},"execution_count":31,"outputs":[]},{"cell_type":"markdown","source":["Plot statistics of the training and testing dataset \n","(`x axis`: digits, `y axis`: number of samples corresponding to the digits)"],"metadata":{"id":"Rx8muKSIrKhe"}},{"cell_type":"code","source":["import numpy as np\n","\n","# Your code\n","digits = np.arange(10)\n","\n","plt.figure(figsize=(10,6))\n","\n","n_digits_train = np.zeros(10)\n","for i in y_train:\n"," n_digits_train[i]+=1\n","\n","plt.bar(digits,n_digits_train) \n","\n","n_digits_test = np.zeros(10)\n","for i in y_test:\n"," n_digits_test[i]+=1\n","\n","plt.bar(digits,n_digits_test) \n","\n","plt.legend([\"Training dataset\", \"Testing dataset\"])\n","# plt.grid(True)"],"metadata":{"id":"37kehTG_6Pi4","executionInfo":{"status":"ok","timestamp":1657696553253,"user_tz":-330,"elapsed":512,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"47aa6e6f-3a24-4001-f891-2d3ddc6f7f04","colab":{"base_uri":"https://localhost:8080/","height":391}},"execution_count":32,"outputs":[{"output_type":"execute_result","data":{"text/plain":[""]},"metadata":{},"execution_count":32},{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["# model building\n","\n","# You are supposed to look at some CNN architectures and add convolutional layers along with MaxPooling, specifying the kernel size, pooling size, activation \n","\n","model = tf.keras.models.Sequential([\n"," tf.keras.layers.Conv2D(32,(3,3), input_shape=(28,28,1)),\n"," tf.keras.layers.Activation('relu'),\n"," tf.keras.layers.MaxPool2D(pool_size=(2,2)),\n"," tf.keras.layers.Flatten(),\n"," tf.keras.layers.Dense(10, activation='softmax')\n","])\n","\n","model.summary()"],"metadata":{"id":"f-bteMURYaJf","executionInfo":{"status":"ok","timestamp":1657696557836,"user_tz":-330,"elapsed":515,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"aad65636-bfe5-4857-9505-e148586a72b8","colab":{"base_uri":"https://localhost:8080/"}},"execution_count":33,"outputs":[{"output_type":"stream","name":"stdout","text":["Model: \"sequential_3\"\n","_________________________________________________________________\n"," Layer (type) Output Shape Param # \n","=================================================================\n"," conv2d_1 (Conv2D) (None, 26, 26, 32) 320 \n"," \n"," activation_1 (Activation) (None, 26, 26, 32) 0 \n"," \n"," max_pooling2d_1 (MaxPooling (None, 13, 13, 32) 0 \n"," 2D) \n"," \n"," flatten_3 (Flatten) (None, 5408) 0 \n"," \n"," dense_7 (Dense) (None, 10) 54090 \n"," \n","=================================================================\n","Total params: 54,410\n","Trainable params: 54,410\n","Non-trainable params: 0\n","_________________________________________________________________\n"]}]},{"cell_type":"code","source":["# Compile the model (add optimizers and metrics)\n","model.compile(\n"," optimizer=tf.keras.optimizers.Adam(0.001),\n"," loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n"," metrics=[tf.keras.metrics.SparseCategoricalAccuracy()],\n",")\n","\n","# Fit the model on the training data (specify validation_split, read about validation if new to you)\n","\n","model.fit(\n"," x_train_norm,\n"," y_train,\n"," epochs=10,\n"," validation_data=(X_val_norm,Y_val),\n",")"],"metadata":{"id":"dvoiYzsUYmFb","executionInfo":{"status":"ok","timestamp":1657696644079,"user_tz":-330,"elapsed":82759,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"831bf75c-a8bc-42dd-bad1-bb0a729a0fb3","colab":{"base_uri":"https://localhost:8080/"}},"execution_count":34,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/10\n","1875/1875 [==============================] - 14s 3ms/step - loss: 0.2077 - sparse_categorical_accuracy: 0.9411 - val_loss: 0.0826 - val_sparse_categorical_accuracy: 0.9755\n","Epoch 2/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0788 - sparse_categorical_accuracy: 0.9771 - val_loss: 0.0719 - val_sparse_categorical_accuracy: 0.9784\n","Epoch 3/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0602 - sparse_categorical_accuracy: 0.9821 - val_loss: 0.0591 - val_sparse_categorical_accuracy: 0.9812\n","Epoch 4/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0495 - sparse_categorical_accuracy: 0.9851 - val_loss: 0.0574 - val_sparse_categorical_accuracy: 0.9799\n","Epoch 5/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0425 - sparse_categorical_accuracy: 0.9872 - val_loss: 0.0533 - val_sparse_categorical_accuracy: 0.9831\n","Epoch 6/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0365 - sparse_categorical_accuracy: 0.9893 - val_loss: 0.0585 - val_sparse_categorical_accuracy: 0.9816\n","Epoch 7/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0303 - sparse_categorical_accuracy: 0.9911 - val_loss: 0.0586 - val_sparse_categorical_accuracy: 0.9825\n","Epoch 8/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0262 - sparse_categorical_accuracy: 0.9921 - val_loss: 0.0621 - val_sparse_categorical_accuracy: 0.9822\n","Epoch 9/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0228 - sparse_categorical_accuracy: 0.9932 - val_loss: 0.0552 - val_sparse_categorical_accuracy: 0.9829\n","Epoch 10/10\n","1875/1875 [==============================] - 5s 3ms/step - loss: 0.0192 - sparse_categorical_accuracy: 0.9941 - val_loss: 0.0556 - val_sparse_categorical_accuracy: 0.9844\n"]},{"output_type":"execute_result","data":{"text/plain":[""]},"metadata":{},"execution_count":34}]},{"cell_type":"markdown","source":["## Model"],"metadata":{"id":"kWlpCWdAr8d3"}},{"cell_type":"code","source":["# Your code\n","p = model.predict(X_test_norm)\n","pred = np.argmax(p,axis=1)\n","\n","fig, axis = plt.subplots(1,15,figsize=(25,5))\n","\n","\n","for i in range(0,15):\n"," axis[i].imshow(X_test[i], cmap='Greys')\n"," \n","for i, ax in enumerate(axis.flat):\n"," ax.set(xlabel= \"Pred: {}\".format(pred[i]))"],"metadata":{"id":"1L07EyQ0Yion","executionInfo":{"status":"ok","timestamp":1657696645853,"user_tz":-330,"elapsed":1787,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"81dbe438-ef70-46eb-dce2-1967f8123c1c","colab":{"base_uri":"https://localhost:8080/","height":142}},"execution_count":35,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["from sklearn.metrics import r2_score\n","\n","accu = r2_score(Y_test,pred)\n","print(\"R2 score: {:.2f}%\".format(accu*100))"],"metadata":{"id":"xv8xM5sPZAwV","executionInfo":{"status":"ok","timestamp":1657696650478,"user_tz":-330,"elapsed":484,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"824a0f11-3a13-46ab-a445-0e172b186084","colab":{"base_uri":"https://localhost:8080/"}},"execution_count":36,"outputs":[{"output_type":"stream","name":"stdout","text":["R2 score: 95.79%\n"]}]},{"cell_type":"markdown","source":["## Predict some images\n","Print the image along with its label (true value) and predicted value."],"metadata":{"id":"ml1Kna_DuJrL"}},{"cell_type":"code","source":["loss, acc = model.evaluate(X_test_norm,Y_test)\n","print(\"Loss: {:.3f}, Accuracy: {:.2f}%\".format(loss,acc*100))"],"metadata":{"id":"qioZul7_uiYq","executionInfo":{"status":"ok","timestamp":1657696654107,"user_tz":-330,"elapsed":833,"user":{"displayName":"Rohit Kumar gupta","userId":"03710792288665421186"}},"outputId":"58e64d0f-6b99-4a2d-8a78-102e5170c548","colab":{"base_uri":"https://localhost:8080/"}},"execution_count":37,"outputs":[{"output_type":"stream","name":"stdout","text":["63/63 [==============================] - 0s 3ms/step - loss: 0.0572 - sparse_categorical_accuracy: 0.9845\n","Loss: 0.057, Accuracy: 98.45%\n"]}]}]} \ No newline at end of file