{"cells":[{"cell_type":"markdown","metadata":{"id":"pVk1SP3EaUvn"},"source":["# Classification with Convolutional Neural Network Using PyTorch\n","\n","ในบทเรียนนี้เราจะทดลองสร้าง Convolutional Neural Network ซึ่งเป็น Neural Network ที่ใช้ฟีเจอร์เชิงพื้นที่ (spatial features) ของภาพเพื่อนำมาช่วยใช้ในการเทรนและทำนายประเภทของภาพ โดยในตัวอย่างนี้เราจะใช้ภาพที่เราได้ทดลองใช้ตอนเรียนเรื่อง Principal Component Analysis (PCA) ซึ่งเป็นภาพที่ผู้ป่วยทดลองวาดนาฬิกา มาทดลองสร้างโมเดลเพื่อแบ่งประเภทกัน\n","\n","ชุดข้อมูล: https://github.com/cccnlab/CDT-API-Network"]},{"cell_type":"markdown","metadata":{},"source":["โดยในขั้น ตอนของการสร้าง Convolutional Neural Network (CNN) นั้น ประกอบไปด้วย\n","1. เตรียม Dataset\n","2. กําหนด input layer โดยมีจํานวนมิติเท่ากับข้อมูลที่เตรียมมา\n","3. กําหนด convolutional layer และ pooling layer โดยอาจทําซํ้าได้หลายรอบสําหรับ\n","feature extraction และการลดขนาดมิติของข้อมูลในแต่ละ layer\n","4. กําหนด fully-connected (dense) layer\n","5. กําหนด logics layer หรือ out put layer ซึ่งในกรณีนี้คือการทํา classification ของ\n","ภาพนาฬิกาที่ถูกวาดโดยผู้ป่วยว่าปกติ (normal) หรือ ผิดปกติ (abnormal)"]},{"cell_type":"markdown","metadata":{},"source":["![Basic-architecture-of-CNN.ppm.png](data:image/png;base64,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)\n","
\n","[ที่มา](https://www.researchgate.net/figure/Basic-architecture-of-CNN_fig3_335086346)"]},{"cell_type":"markdown","metadata":{},"source":["Convolution คือการดําเนินการทางคณิตศาสตร์ $f * g$ เพื่อแสดงให้เห็นผลของฟังก์ชัน $f$ ที่ เปลี่ยนแปลงไปเมื่อมีฟังก์ชัน $g$ โดยในรูปด้านล่าง สีม่วงคือ $f = [6,2,5,2,1,0,4,3,6,4]$ สีเขียว คือ $g = [6,5,4]$ และสีแดงคือ $f * g = [66,45,44,17,22,32,63,64]$"]},{"cell_type":"markdown","metadata":{},"source":["![Screenshot_2566-07-13_at_20.56.03.png](data:image/png;base64,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)"]},{"cell_type":"markdown","metadata":{},"source":["เพื่อเปรียบเทียบให้เห็นภาพชัด Convolution ถูกนํามาใช้ใน classic machine learning อย่างมาก สําหรับ การทํา feature extraction ในรูปแบบ 2D โดยมี $f$ คือ image และ $g$ คือ\n","filter (หรืออาจเรียกว่า kernel) โดยใน convolutional layer สามารถเลือกปรับ ขนาดของ input (padding) หรือ ขนาดการเลื่อนของการแสกน (stride) ได้ มีผลลัพธ์ท่ีได้จากการ คํานวณเรียกว่า Feature Map\n"]},{"cell_type":"markdown","metadata":{},"source":["![1Fw-ehcNBR9byHtho-Rxbtw.png](data:image/png;base64,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)"]},{"cell_type":"markdown","metadata":{},"source":["ในกรณี Back propagation จะมีกระบวนการเรียนรู้เพื่อปรับ filter (kernel) ของ convolutionallayerเพื่อให้มีผลลัพธ์ที่สูงที่สุด\n","โดย convolutional layer จะถูกส่งต่อไปยัง pooling layer เพื่อปรับขนาดภาพ และลดมิติของfeaturemapลง เพื่อให้โมเดลไม่จําเป็นต้องเรียนรู้ค่า weight ที่มีจํานวนมากเกินไป โดยมีเทคนิคที่นิยมคือ max pooling และ average pooling\n","โดย pooling layer สุดท้ายจะถูกรวม และเปลี่ยนรูปไปเป็น fully-connected layer เพื่อคํานวนเป็นoutputlayerใช้สําหรับ classification ภาพนาฬิกาที่ได้ยกนํามาเป็นตัว อย่างในบท นี้"]},{"cell_type":"markdown","metadata":{},"source":["## Download the Data"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":1155,"status":"ok","timestamp":1688465945839,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"PMI86_jqaUvq","outputId":"646ab883-3dcd-4e8e-e857-1c90d3c07305"},"outputs":[{"name":"stdout","output_type":"stream","text":["--2023-07-04 10:19:02-- https://github.com/ichatnun/brainCodeCamp/raw/main/Fundamentals/DimensionalityReduction/data/CDT/clock_images.pickle\n","Resolving github.com (github.com)... 140.82.112.3\n","Connecting to github.com (github.com)|140.82.112.3|:443... connected.\n","HTTP request sent, awaiting response... 302 Found\n","Location: https://raw.githubusercontent.com/ichatnun/brainCodeCamp/main/Fundamentals/DimensionalityReduction/data/CDT/clock_images.pickle [following]\n","--2023-07-04 10:19:03-- https://raw.githubusercontent.com/ichatnun/brainCodeCamp/main/Fundamentals/DimensionalityReduction/data/CDT/clock_images.pickle\n","Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.111.133, 185.199.109.133, 185.199.110.133, ...\n","Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.111.133|:443... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 1106083 (1.1M) [application/octet-stream]\n","Saving to: ‘images.pickle’\n","\n","images.pickle 100%[===================>] 1.05M --.-KB/s in 0.05s \n","\n","2023-07-04 10:19:03 (20.2 MB/s) - ‘images.pickle’ saved [1106083/1106083]\n","\n","--2023-07-04 10:19:03-- https://github.com/ichatnun/brainCodeCamp/raw/main/Fundamentals/DimensionalityReduction/data/CDT/labels.pickle\n","Resolving github.com (github.com)... 140.82.113.4\n","Connecting to github.com (github.com)|140.82.113.4|:443... connected.\n","HTTP request sent, awaiting response... 302 Found\n","Location: https://raw.githubusercontent.com/ichatnun/brainCodeCamp/main/Fundamentals/DimensionalityReduction/data/CDT/labels.pickle [following]\n","--2023-07-04 10:19:03-- https://raw.githubusercontent.com/ichatnun/brainCodeCamp/main/Fundamentals/DimensionalityReduction/data/CDT/labels.pickle\n","Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.108.133, 185.199.109.133, 185.199.110.133, ...\n","Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.108.133|:443... connected.\n","HTTP request sent, awaiting response... 200 OK\n","Length: 630 [application/octet-stream]\n","Saving to: ‘labels.pickle’\n","\n","labels.pickle 100%[===================>] 630 --.-KB/s in 0s \n","\n","2023-07-04 10:19:04 (55.1 MB/s) - ‘labels.pickle’ saved [630/630]\n","\n"]}],"source":["# Download the dataset which contrains images and labels of the Clock Drawing Test (CDT)\n","\n","!wget -O images.pickle https://github.com/ichatnun/brainCodeCamp2023/raw/main/Fundamentals/DimensionalityReduction/data/CDT/clock_images.pickle\n","!wget -O labels.pickle https://github.com/ichatnun/brainCodeCamp2023/raw/main/Fundamentals/DimensionalityReduction/data/CDT/labels.pickle"]},{"cell_type":"code","execution_count":1,"metadata":{"id":"k-ILPP8ZaUvt"},"outputs":[],"source":["# import libraries\n","import torch\n","import torch.nn as nn\n","import torch.nn.functional as F\n","import torch.optim as optim # module ที่เก็บ optimizer ต่างๆ ที่เราจะเอามาใช้ในการ train model\n","from torch.utils.data import Dataset, DataLoader # โหลด dataset มาใช้งาน\n","import torchvision # module สำหรับ computer vision ของ PyTorch\n","import torchvision.transforms as transforms # module สำหรับ transform รูปภาพ เช่น การปรับขนาดรูปภาพ การทำ data augmentation\n","from sklearn.model_selection import train_test_split # ใช้สำหรับแบ่งข้อมูลเป็น train/validation/test set\n","from glob import glob # ใช้สำหรับหา path ของไฟล์ในโฟลเดอร์\n","from PIL import Image # ใช้สำหรับโหลดรูปภาพ และแปลงรูปภาพให้เป็น array\n","from matplotlib import pyplot as plt # ใช้สำหรับ plot รูปภาพ\n","import pickle # ใช้สำหรับ save และ load ข้อมูล\n","import pandas as pd # ใช้สำหรับจัดการข้อมูลในรูปแบบของตาราง\n","import numpy as np # ใช้สำหรับจัดการข้อมูลในรูปแบบของ array\n","import os # ใช้สำหรับจัดการไฟล์และโฟลเดอร์"]},{"cell_type":"markdown","metadata":{"id":"mVUmiMvDaUvt"},"source":["## Prepare Data\n","\n","การเตรียมข้อมูลทำได้โดยการอ่านข้อมูลภาพจาก `images.pickle` และประเภทของภาพจาก `labels.pickle` เพื่อใช้ในการเตรียมข้อมูล จากนั้นเปลี่ยนเป็น dataframe"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"vDNsj1yjaUvu"},"outputs":[],"source":["# โหลดข้อมูลไฟล์ .pickle ที่เราได้ทำการ save ไว้ก่อนหน้านี้ และโหลดมาผ่าน wget จาก github\n","with open(\"images.pickle\", \"rb\") as handle:\n"," images = pickle.load(handle)\n","with open(\"labels.pickle\", \"rb\") as handle:\n"," labels = pickle.load(handle)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"C0CyAxm6aUvu"},"outputs":[],"source":["# แปลงรูปภาพให้เป็น RGB array และเก็บไว้ใน list\n","list_images = []\n","for image in images:\n"," image = Image.fromarray(image) # Convert array to PIL image\n"," image = image.convert(\"RGB\") # Convert to RGB mode\n"," list_images.append(image)"]},{"cell_type":"markdown","metadata":{"id":"65F202JBaUvv"},"source":["สร้าง dataframe ที่เก็บข้อมูลที่จำเป็นต้องใช้ โดยมี column ดังนี้\n","- `Images` : ภาพที่ใช้ train\n","- `encoded_labels` : ค่า label โดย 0 คือ ปกติ และ 1 คือ ผิดปกติ"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":656,"status":"ok","timestamp":1688465954893,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"FTOeHGI-aUvv","outputId":"8fc4a0b1-f088-44fa-8ba3-69862da901a5"},"outputs":[{"name":"stderr","output_type":"stream","text":[":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n",":4: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)\n"]}],"source":["df_data = pd.DataFrame(columns=[\"Images\", \"encoded_labels\"]) # สร้าง dataframe เพื่อเก็บข้อมูลรูปภาพและ label ที่เป็นตัวเลข\n","\n","# นำข้อมูลรูปภาพและ label มาเก็บไว้ใน dataframe\n","for img, label in zip(list_images, labels):\n"," df_data = df_data.append({\"Images\": img, \"encoded_labels\": label}, ignore_index=True)"]},{"cell_type":"markdown","metadata":{"id":"gPyZ2YYJaUvv"},"source":["## Split Data and Prepare DataLoaders\n","\n","แบ่งข้อมูลเป็น train set (90%) และ test set (10%) โดยใช้ `train_test_split` จาก `sklearn.model_selection`"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":20,"status":"ok","timestamp":1688465954894,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"NGTw-NzUaUvw","outputId":"fde46b8f-87c0-42d1-a11d-4f991115a838"},"outputs":[{"data":{"text/plain":["(54, 6)"]},"execution_count":6,"metadata":{},"output_type":"execute_result"}],"source":["# แบ่งข้อมูลเป็น train/validation/test set โดยใช้ train/validation set อยู่ในสัดส่วน 90:10\n","train_data, test_data = train_test_split(df_data, test_size=0.1, random_state=42)\n","len(train_data), len(test_data)"]},{"cell_type":"markdown","metadata":{"id":"roaNKSnnaUvw"},"source":["สร้างชุดข้อมูล dataset จาก training และ testing โดยใช้ `torch.utils.data.Dataset` and `torch.utils.data.DataLoader` โดย\n","\n","- `Dataset` เป็น class ที่ใช้ในการกำหนดชุดข้อมูลของเรา โดยปกติจะประกอบด้วย 3 ฟังก์ชันหลักคือ `__init__` ใช้เพื่อประกาศค่าต่างๆ, `__len__` เป็นฟังก์ชันที่ให้ขนาดของชุดข้อมูล, และ `__getitem__` ที่มี index เป็น input และให้ `img` และ `label` ของ index นั้นๆออกมา\n","- `DataLoader` ส่วนมากจะใช้ร่วมกับ `Dataset` ที่เราสร้างขึ้น โดยเมื่อเราประกาศ `Dataset` แล้วจะสามารถนำ `DataLoader` มาครอบ `Dataset` เพื่อใช้ในการดึงข้อมูลออกมาเป็น batch ได้"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"hvHga8XtaUvw"},"outputs":[],"source":["# สร้างขั้นตอนสำหรับการแปลงข้อมูลรูปภาพรวมเป็นตัวแปร transform โดย\n","# 1. ทำการแปลงรูปภาพให้เป็น tensor\n","# 2. ทำการ normalize รูปภาพด้วย mean และ standard deviation ของชุดข้อมูล\n","transform = transforms.Compose(\n"," [transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]\n",")\n","\n","base_path = \"CDT_dataset\" # กำหนด path ของ dataset ที่เราจะเก็บไว้\n","\n","# สร้าง class สำหรับการโหลด dataset ที่เราสร้างขึ้นมา\n","class CDT_dataset(Dataset): \n"," def __init__(self, img_data, base_path, transform):\n"," self.img_data = img_data\n"," self.base_path = base_path\n"," self.transform = transform\n","\n"," def __len__(self): # สร้าง function สำหรับ return จำนวนข้อมูลใน dataset ด้วยการใช้ len(...)\n"," return len(self.img_data)\n","\n"," def __getitem__(self, index): # สร้าง function สำหรับ return ข้อมูลและ label ของข้อมูลใน dataset ด้วยการใช้ index\n"," # โหลดรูปภาพจาก path ที่เก็บไว้\n"," img = self.img_data.iloc[index][\"Images\"]\n"," # ลดขนาดภาพให้มีขนาด 32x32 เพื่อใช้ในการทำนาย\n"," img = img.resize((32, 32)) \n"," # ถ้ามีการสร้างขั้นตอนการแปลงข้อมูลรูปภาพ ให้ทำการแปลงข้อมูลรูปภาพด้วย\n"," if self.transform is not None:\n"," img = self.transform(img)\n"," # สร้าง label ของข้อมูล\n"," label = self.img_data.iloc[index][\"encoded_labels\"]\n"," # ส่งข้อมูลรูปภาพและ label กลับไป\n"," return img, label\n","\n","\n","train_dataset = CDT_dataset(train_data, base_path, transform) # สร้าง train dataset\n","test_dataset = CDT_dataset(test_data, base_path, transform) # สร้าง test dataset"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"Zg0tWa2jaUvx"},"outputs":[],"source":["# สร้าง train loader และ test loader โดยกำหนดให้มี batch size เท่ากับ 4 และ shuffle ข้อมูล\n","# โดยที่ DataLoader คือ class สำหรับการโหลดข้อมูลในรูปแบบของ batch\n","train_loader = DataLoader(train_dataset, batch_size=4, shuffle=True) # สร้าง train loader\n","test_loader = DataLoader(test_dataset, batch_size=4, shuffle=True) # สร้าง test loader"]},{"cell_type":"markdown","metadata":{"id":"jiBIfq7haUvx"},"source":["เขียนฟังก์ชันเพื่อยกตัวอย่างภาพที่ใช้ train โดยใช้ `next(iter(train_loader))` เพื่อดึงตัวอย่างออกมา 1 batch ในด้านบนเรากำหนดให้ `batch_size` มีค่าเท่ากับ 4"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":210},"executionInfo":{"elapsed":12,"status":"ok","timestamp":1688465954897,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"BkCBEg1daUvx","outputId":"bf4792c8-161d-430b-a27d-4c6f0f140f12"},"outputs":[{"data":{"image/png":"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","text/plain":["
"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["abnormal normal abnormal abnormal\n"]}],"source":["# สร้างฟังก์ชันสำหรับการแสดงรูปภาพและ label ของข้อมูลใน dataset\n","def imshow(img):\n"," img = img / 2 + 0.5\n"," npimg = img.numpy()\n"," plt.imshow(np.transpose(npimg, (1, 2, 0)))\n"," plt.show()\n","\n","images, labels = next(iter(train_loader)) # ดึงข้อมูลจาก train loader มาเก็บไว้ในตัวแปร images และ labels\n","# iter(...) คือ function สำหรับการสร้าง iterator ของ train loader\n","# next(...) คือ function สำหรับการดึงข้อมูลจาก iterator ของ train loader มาเก็บไว้ในตัวแปร images และ labels\n","\n","SD_type = {0: \"normal\", 1: \"abnormal\"} # สร้าง dictionary สำหรับแปลง label ที่เป็นตัวเลขเป็นชื่อของ class\n","\n","imshow(torchvision.utils.make_grid(images)) # แสดงรูปภาพที่อยู่ในตัวแปร images\n","classes = (\"normal\", \"abnormal\") # กำหนดชื่อของ class ที่เรามี\n","print(\" \".join(f\"{classes[labels[j]]:5s}\" for j in range(4))) # แสดง label ของข้อมูลที่อยู่ในตัวแปร labels"]},{"cell_type":"markdown","metadata":{"id":"TSx2h4KHaUvx"},"source":["## Introduction to Convolution\n","convolution เป็นวิธีการนึ่งที่ที่ดูการเปลี่ยนแปลงของฟังก์ชั่นหนึ่ง เมื่อมีอีกฟังก์ชั่นเข้ามา ใน deep-learning convolution ถูกนำมาใช้อย่างแพร่หลายในงานที่เกี่ยวกับรูปภาพเพื่อดึงข้อมูลเชิงพื้นที่ (spatial feature) ออกมาจากรูปภาพ โดย pytorch จะใช้คำสั่ง `nn.Conv2d` ในการทำ convolution\n","โดย parameters ที่สำคัญจะประกอบไปด้วย\n","* `in_channels`: input image channel\n","* `out_channels`: output channel หลังจาก convolve แล้ว\n","* `kernel_size`: ขนาดของ filter ที่นำมา convolve กับรูปของเรา\n","* `stride`: filter จะเคลื่อนที่ทีละกี่ pixel ค่า default = 1\n","* `padding`: padding รูปภาพ ค่า default = 0\n","\n","ภาพที่ได้หลังจากการ convolution อาจจะมีขนาด/depthเปลี่ยนไปขึ้นอยู่กับ parametersซึ่งเราสามารถคำนวนขนาดของภาพที่ออกมาได้ดังนี้\n","\n","`output_size = ((input_size - kernel_size + 2 * padding) / stride) + 1`"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":394,"status":"ok","timestamp":1688465955282,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"_czkGldbaUvy","outputId":"30328401-fbb4-4a1a-dcb3-363500d24efd"},"outputs":[{"data":{"text/plain":["torch.Size([1, 3, 5, 5])"]},"execution_count":10,"metadata":{},"output_type":"execute_result"}],"source":["# สมมุติว่ามีรูปภาพ 1 ภาพ ขนาด 5x5 และมี 3 ช่องสี\n","input = torch.randn(1, 3, 5, 5)\n","input.shape"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":27,"status":"ok","timestamp":1688465955297,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"n-UVz5IuaUvy","outputId":"20bd6819-c464-4c54-d3ed-83aa981651cd"},"outputs":[{"data":{"text/plain":["torch.Size([1, 16, 3, 3])"]},"execution_count":11,"metadata":{},"output_type":"execute_result"}],"source":["conv = nn.Conv2d(in_channels=3, # จำนวนช่องสีของภาพ\n"," out_channels=16, # จำนวนช่องสีของภาพที่เราต้องการให้เป็น output\n"," kernel_size=3, # ขนาดของ filter\n"," stride=1, # ขนาดของการเคลื่อนที่ของ filter\n"," padding=0) # จำนวนช่องสีที่เราต้องการให้เพิ่มเติมเพื่อให้ขนาดของภาพเท่าเดิม\n","# output size = (5 - 3 + 2*0)/1 + 1 = 3\n","output = conv(input) # ทำการคำนวนด้วยการป้อน input ลงไปใน conv ซึ่งเป็น object ที่เราสร้างไว้ให้ทำหน้าที่เป็น convolution ในสองมิติ \n","output.shape # depth ของภาพจะเพิ่มเป็น 16และ dimension ของภาพจะเท่ากับการคำนวนคือ 3"]},{"cell_type":"markdown","metadata":{"id":"XkwGhz4xaUvy"},"source":["## Train the Model\n","\n","สร้าง model โดยใช้ `torch.nn.Module` ซึ่งเป็นตัวแบบของ Neural Network model ใน pytorch โดยมี layer ดังนี้\n","- Convolutional layer (`nn.Conv2d`) ซึ่งจะรับค่าทั้งหมด 3 ค่าคือ input depth, output depth และ kernel size ยกตัวอย่างเช่นในเลเยอร์แรก `nn.Conv2d(3, 6, 5)` จะรับ input depth ของภาพเท่ากับ 3 นั่นก็คือ RGB channel นั่นเอง และ output ออกมาเป็นภาพที่มี depth เท่ากับ 6\n","- จากนั้นก็จะส่งต่อไปให้ `nn.MaxPool2d(2, 2)` ซึ่งมี kernel size เท่ากับ 2 และ stride เท่ากับ 2 โดยจะลดขนาดของภาพเป็นครึ่งหนึ่งนั่นเอง (แต่ depth เท่าเดิม)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"EDT0xGwNaUvy"},"outputs":[],"source":["class Net(nn.Module): # สร้าง class สำหรับการสร้างโมเดลจาก nn.Module\n"," def __init__(self): # สร้าง constructor ของ class\n"," super().__init__() # เรียกใช้ constructor ของ nn.Module\n"," self.conv1 = nn.Conv2d(3, 6, 5) # สร้าง convolution layer แรก โดยมี in_channels เท่ากับ 3 และ out_channels เท่ากับ 6 และมี kernel_size เท่ากับ 5\n"," self.pool = nn.MaxPool2d(2, 2) # สร้าง pooling layer โดยมี kernel_size เท่ากับ 2 และ stride เท่ากับ 2\n"," self.conv2 = nn.Conv2d(6, 16, 5) # สร้าง convolution layer ที่สอง โดยมี in_channels เท่ากับ 6 และ out_channels เท่ากับ 16 และมี kernel_size เท่ากับ 5\n"," self.fc1 = nn.Linear(16 * 5 * 5, 120) # สร้าง fully connected layer แรก โดยมี in_features เท่ากับ 16 * 5 * 5 และ out_features เท่ากับ 120\n"," self.fc2 = nn.Linear(120, 84) # สร้าง fully connected layer ที่สอง โดยมี in_features เท่ากับ 120 และ out_features เท่ากับ 84\n"," self.fc3 = nn.Linear(84, 10) # สร้าง fully connected layer ที่สาม โดยมี in_features เท่ากับ 84 และ out_features เท่ากับ 10\n","\n"," def forward(self, x): # สร้าง method สำหรับการทำ forward propagation\n"," x = self.pool(F.relu(self.conv1(x))) # ทำการ convolution และ pooling และ activation function (ReLu) ใน convolution layer แรก\n"," x = self.pool(F.relu(self.conv2(x))) # ทำการ convolution และ pooling และ activation function (ReLu) ใน convolution layer ที่สอง\n"," x = torch.flatten(x, 1) # ทำการ flatten ข้อมูลใน x ให้เหลือแค่ batch dimension\n"," x = F.relu(self.fc1(x)) # ทำการแปลงผ่าน fully connected และ activation function (ReLu) ใน fully connected layer แรก\n"," x = F.relu(self.fc2(x)) # ทำการแปลงผ่าน fully connected และ activation function (ReLu) ใน fully connected layer ที่สอง\n"," x = self.fc3(x) # ทำการแปลงผ่าน fully connected ใน fully connected layer สุดท้าย\n"," return x # ส่งค่า x กลับ\n","\n","\n","model = Net() # สร้าง object ของ class Net"]},{"cell_type":"markdown","metadata":{},"source":["**เราทราบได้อย่างไรว่า class ที่เราสร้างขึ้น มี network layer เชื่อมต่อกันอย่างไร**\n","
\n"," \n"," เฉลย\n"," \n"," เราสามารถดูได้จากขั้นตอนการแปลง input ผ่าน method `forward` ที่อยู่ใน class ได้ โดยจากตัวอย่างด้านบน จะพบว่า network ของเรามีการเชื่อมต่อกันแบบเรียงลำดับดังนี้\n","
    \n","
  1. conv1 + ReLu
  2. \n","
  3. conv2 + ReLu
  4. \n","
  5. fc1 + ReLu
  6. \n","
  7. fc2 + ReLu
  8. \n","
  9. fc3
  10. \n","
\n"," โดยที่ activation function นั้นจะมีหรือไม่มีหลังจากที่ทำการคำนวณใน layer แต่ละอันก็ได้\n","
"]},{"cell_type":"markdown","metadata":{"id":"yL6TjxkXaUvy"},"source":["กำหนด `hyperparameters` ที่จำเป็นต้องใช้ ได้แก่ loss function ซึ่งคือ `CrossEntropyLoss`, optimizer ซึ่งคือ `Adam` ที่รับพารามิเตอร์ของโมเดล จากนั้นเทรนโมเดลโดยใส่ชุดข้อมูลเข้าไปในโมเดล และทำการอัพเดทโมเดลด้วยการคำนวณ gradient (`loss.backward()`) และอัพเดทพารามิเตอร์ของโมเดล (`optimizer.step()`)"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"TsUTo5x0aUvz"},"outputs":[],"source":["# Define a Loss function and optimizer\n","\n","criterion = nn.CrossEntropyLoss() # กำหนด loss function ที่ใช้ในการคำนวนค่า loss เป็นค่า Cross Entropy Loss ซึ่งหมายถึง การคำนวนค่า loss โดยใช้ค่าความแตกต่างของค่าความน่าจะเป็นของคลาสที่เราต้องการกับค่าความน่าจะเป็นของคลาสอื่นๆ\n","optimizer = optim.Adam(model.parameters(), lr=0.001) # กำหนด optimizer ที่ใช้ในการปรับค่า parameter ของโมเดลให้เป็น optimizer แบบ Adam ซึ่งเป็น optimizer ที่ใช้ในการปรับค่า parameter โดยใช้การคำนวนค่า gradient ของ parameter แต่ละตัวเพื่อใช้ในการปรับค่า parameter โดยใช้ learning rate เท่ากับ 0.001"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":14006,"status":"ok","timestamp":1688465969279,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"4tXoLWriaUvz","outputId":"5791df0c-64e4-41ba-971b-ae9caa12876a"},"outputs":[{"name":"stdout","output_type":"stream","text":["[epoch: 1 ] loss: 20.075\n","[epoch: 2 ] loss: 8.571\n","[epoch: 3 ] loss: 9.308\n","[epoch: 4 ] loss: 7.529\n","[epoch: 5 ] loss: 6.981\n","[epoch: 6 ] loss: 7.147\n","[epoch: 7 ] loss: 6.895\n","[epoch: 8 ] loss: 7.352\n","[epoch: 9 ] loss: 6.986\n","[epoch: 10 ] loss: 7.029\n","[epoch: 11 ] loss: 7.027\n","[epoch: 12 ] loss: 7.004\n","[epoch: 13 ] loss: 7.241\n","[epoch: 14 ] loss: 7.126\n","[epoch: 15 ] loss: 6.847\n","[epoch: 16 ] loss: 6.681\n","[epoch: 17 ] loss: 6.850\n","[epoch: 18 ] loss: 6.481\n","[epoch: 19 ] loss: 6.609\n","[epoch: 20 ] loss: 5.753\n","[epoch: 21 ] loss: 5.363\n","[epoch: 22 ] loss: 3.350\n","[epoch: 23 ] loss: 3.134\n","[epoch: 24 ] loss: 1.701\n","[epoch: 25 ] loss: 2.908\n","[epoch: 26 ] loss: 1.856\n","[epoch: 27 ] loss: 1.559\n","[epoch: 28 ] loss: 0.711\n","[epoch: 29 ] loss: 0.615\n","[epoch: 30 ] loss: 0.605\n","[epoch: 31 ] loss: 1.207\n","[epoch: 32 ] loss: 0.896\n","[epoch: 33 ] loss: 0.525\n","[epoch: 34 ] loss: 0.756\n","[epoch: 35 ] loss: 0.229\n","[epoch: 36 ] loss: 0.151\n","[epoch: 37 ] loss: 0.324\n","[epoch: 38 ] loss: 0.378\n","[epoch: 39 ] loss: 0.846\n","[epoch: 40 ] loss: 0.418\n","[epoch: 41 ] loss: 0.150\n","[epoch: 42 ] loss: 0.042\n","[epoch: 43 ] loss: 0.023\n","[epoch: 44 ] loss: 0.033\n","[epoch: 45 ] loss: 0.032\n","[epoch: 46 ] loss: 0.022\n","[epoch: 47 ] loss: 0.028\n","[epoch: 48 ] loss: 0.022\n","[epoch: 49 ] loss: 0.017\n","[epoch: 50 ] loss: 0.021\n","[epoch: 51 ] loss: 0.014\n","[epoch: 52 ] loss: 0.017\n","[epoch: 53 ] loss: 0.018\n","[epoch: 54 ] loss: 0.011\n","[epoch: 55 ] loss: 0.014\n","[epoch: 56 ] loss: 0.014\n","[epoch: 57 ] loss: 0.007\n","[epoch: 58 ] loss: 0.013\n","[epoch: 59 ] loss: 0.008\n","[epoch: 60 ] loss: 0.011\n","[epoch: 61 ] loss: 0.006\n","[epoch: 62 ] loss: 0.011\n","[epoch: 63 ] loss: 0.007\n","[epoch: 64 ] loss: 0.010\n","[epoch: 65 ] loss: 0.006\n","[epoch: 66 ] loss: 0.006\n","[epoch: 67 ] loss: 0.007\n","[epoch: 68 ] loss: 0.008\n","[epoch: 69 ] loss: 0.007\n","[epoch: 70 ] loss: 0.007\n","[epoch: 71 ] loss: 0.007\n","[epoch: 72 ] loss: 0.007\n","[epoch: 73 ] loss: 0.007\n","[epoch: 74 ] loss: 0.002\n","[epoch: 75 ] loss: 0.002\n","[epoch: 76 ] loss: 0.006\n","[epoch: 77 ] loss: 0.004\n","[epoch: 78 ] loss: 0.006\n","[epoch: 79 ] loss: 0.005\n","[epoch: 80 ] loss: 0.005\n","[epoch: 81 ] loss: 0.005\n","[epoch: 82 ] loss: 0.003\n","[epoch: 83 ] loss: 0.003\n","[epoch: 84 ] loss: 0.003\n","[epoch: 85 ] loss: 0.003\n","[epoch: 86 ] loss: 0.002\n","[epoch: 87 ] loss: 0.003\n","[epoch: 88 ] loss: 0.002\n","[epoch: 89 ] loss: 0.004\n","[epoch: 90 ] loss: 0.002\n","[epoch: 91 ] loss: 0.003\n","[epoch: 92 ] loss: 0.002\n","[epoch: 93 ] loss: 0.003\n","[epoch: 94 ] loss: 0.003\n","[epoch: 95 ] loss: 0.003\n","[epoch: 96 ] loss: 0.003\n","[epoch: 97 ] loss: 0.002\n","[epoch: 98 ] loss: 0.001\n","[epoch: 99 ] loss: 0.002\n","[epoch: 100 ] loss: 0.002\n","Finished Training\n"]}],"source":["for epoch in range(100): # ทำการ train model โดยใช้จำนวน epoch เท่ากับ 100\n"," running_loss = 0.0 # กำหนดค่าเริ่มต้นของ running loss เท่ากับ 0\n"," for i, data in enumerate(train_loader, 0): # ทำการวน loop ในแต่ละ batch ของ train_loader\n"," inputs, labels = data # ดึงข้อมูลและ label จาก data\n"," optimizer.zero_grad() # ทำการปรับค่า gradient เป็น 0 ก่อนการคำนวนค่า gradient ของ parameter แต่ละตัว เนื่องจากการคำนวณ gradient ด้วย backpropagation จะเก็บค่า gradient ไว้ในตัวแปรเดียวกันเรื่อยๆ จนกว่าจะมีการเรียกใช้ method zero_grad() เพื่อทำการเคลียร์ค่า gradient ในตัวแปรนั้น\n"," outputs = model(inputs) # ทำการ forward propagation โดยใช้ข้อมูลใน inputs ที่ได้จาก data ที่วน loop อยู่\n"," loss = criterion(outputs, labels) # ทำการคำนวณค่า loss โดยใช้ค่า output ที่ได้จากการทำ forward propagation กับ label ที่ได้จาก data ที่วน loop อยู่\n"," loss.backward() # ทำการคำนวณค่า gradient ของ parameter แต่ละตัวโดยใช้ backpropagation\n"," optimizer.step() # ทำการปรับค่า parameter ของโมเดลด้วยค่า gradient ที่คำนวณได้\n"," running_loss += loss.item() # ทำการบวกค่า loss ในแต่ละ batch เข้ากับ running loss เพื่อใช้ในการคำนวณค่า loss ในแต่ละ epoch\n"," if i % 10 == 9: # ทำการ print ค่า loss ในแต่ละ batch ทุกๆ 10 batch\n"," print(f\"[epoch: {epoch + 1} ] loss: {running_loss / 1:.3f}\")\n"," running_loss = 0.0 # ทำการเคลียร์ค่า running loss เพื่อใช้ในการคำนวณค่า loss ใน batch ถัดไป\n","\n","print(\"Finished Training\")"]},{"cell_type":"code","execution_count":null,"metadata":{"id":"iJWp-Za2aUvz"},"outputs":[],"source":["# Save the model\n","PATH = \"./CDT_model.pth\"\n","torch.save(model.state_dict(), PATH) # ทำการบันทึกโมเดลที่ได้จากการ train เป็นไฟล์ .pth โดย .state_dict() จะเป็นการบันทึกเฉพาะค่า parameter ของโมเดลเท่านั้น"]},{"cell_type":"markdown","metadata":{"id":"jx6txDwxaUvz"},"source":["### Test the Data\n","\n","ทดลองนำภาพตัวอย่างจาก test set มาทำนายด้วยโมเดลที่เราได้เทรนไป"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":210},"executionInfo":{"elapsed":12,"status":"ok","timestamp":1688465969280,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"6jP-6W-PaUvz","outputId":"d3fb40c2-679d-485d-e135-42d6abb5dfae"},"outputs":[{"data":{"image/png":"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","text/plain":["
"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["GroundTruth: normal normal normal abnormal\n"]}],"source":["dataiter = iter(test_loader) # ทำการสร้าง object ของ test_loader ให้เป็น iterator\n","images, labels = next(dataiter) # ทำการดึงข้อมูลและ label จาก test_loader ที่เป็น iterator ออกมา\n","\n","# print images\n","imshow(torchvision.utils.make_grid(images)) # ทำการแสดงรูปภาพที่อยู่ใน images ที่เป็น tensor ในรูปแบบของ grid\n","print(\"GroundTruth: \", \" \".join(f\"{classes[labels[j]]:5s}\" for j in range(4))) # ทำการแสดง label ของรูปภาพที่อยู่ใน images ที่เป็น tensor ในรูปแบบของ grid"]},{"cell_type":"markdown","metadata":{"id":"sn5_zWONaUv0"},"source":["ในการใช้งานจริงเราสามารถสร้างโมเดล และอ่าน weights (หรือ parameters) ของโมเดลที่เทรนไว้แล้ว เพื่อนำมาใช้ทำนายเลยได้ ไม่จำเป็นต้องเทรนใหม่"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":11,"status":"ok","timestamp":1688465969281,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"pO30INrfaUv0","outputId":"608006a4-17fd-413d-d265-157227078645"},"outputs":[{"data":{"text/plain":[""]},"execution_count":17,"metadata":{},"output_type":"execute_result"}],"source":["model = Net() # สร้าง object ของ class Net ที่เป็นโมเดลที่เราสร้างไว้\n","model.load_state_dict(torch.load(PATH)) # ทำการโหลดค่า parameter ของโมเดลจากไฟล์ .pth ที่เราบันทึกไว้ก่อนหน้านี้"]},{"cell_type":"code","execution_count":null,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":8,"status":"ok","timestamp":1688465969281,"user":{"displayName":"Korrawiz Chotayapa","userId":"12619977060284176060"},"user_tz":-420},"id":"i-5-5gRmaUv0","outputId":"27c983dc-ed0a-4120-8480-47d24634dd80"},"outputs":[{"name":"stdout","output_type":"stream","text":["Predicted: normal normal normal abnormal\n"]}],"source":["# Predictions\n","outputs = model(images)\n","\n","_, predicted = torch.max(outputs, 1)\n","\n","print(\"Predicted: \", \" \".join(f\"{classes[predicted[j]]:5s}\" for j in range(4)))"]},{"cell_type":"markdown","metadata":{"id":"WK_9iPgLaUv2"},"source":["ในตัวอย่างนี้เราเพียงแค่ต้องการโชว์กระบวนการเบื้องต้นของการสร้างและเทรนโมเดล Convolutional Neural Network เท่านั่นแต่ในเทรนโมเดลจริงนั่นเรายังต้องคำนึงถึงปัจจัยอื่นๆอีกมากมาย เช่น การ overfit ของโมเดล, การปรับหน้าตาของโมเดล, การใช้ optimizer แบบอื่นๆ ที่เป็นปัจจัยในการเทรนโมเดลให้มีประสิทธิภาพ\n","\n","นอกจากนี้ปัจจุบัน CNN ได้ถูกออกแบบและพัฒนาให้มีรูปแบบที่ซับซ้อนขึ้นเพื่อให้ได้ความแม่นยำที่มากขึ้นเช่น VGG, DenseNet, ResNet รวมถึงจะมีการเทรนโมเดลเหล่านั้นกับข้อมูลขนาดใหญ่แล้วค่อยนำมา fine-tune เพื่อให้เหมาะสมกับ task ที่เราต้องการทำให้ลดเวลาในการเทรนและเพิ่มความแม่นยำให้กับโมเดลได้ โดยในบทถัดๆไปจะเป็นการยกตัวอย่างการ download โมเดลที่ซับซ้อนขึ้นที่ถูกเทรนมาแล้ว และนำมาใช้ในงานของเรา เช่น\n","* Classification\n","* Object detection\n","* Semantic Segmentation\n","\n","

\n","**ผู้จัดเตรียม code ใน tutorial**: ดร. ฐิติพัทธ อัชชะกุลวิสุทธิ์"]},{"cell_type":"markdown","metadata":{},"source":[]}],"metadata":{"accelerator":"GPU","colab":{"gpuType":"T4","provenance":[]},"kernelspec":{"display_name":"Python 3","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.11"}},"nbformat":4,"nbformat_minor":0}