{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":3004,"databundleVersionId":861823,"sourceType":"competition"}],"dockerImageVersionId":30587,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# Phần 1: Nhập các thư viện cần thiết\n------","metadata":{"_uuid":"b0093541-188e-43e9-acb2-d917ae3a5010","_cell_guid":"cadbfcb2-82c3-470f-bad6-eade54542125","trusted":true}},{"cell_type":"markdown","source":"Cơ sở dữ liệu MNIST (Cơ sở dữ liệu của Viện Tiêu chuẩn và Công nghệ Quốc gia đã sửa đổi) là cơ sở dữ liệu lớn gồm các chữ số viết tay thường được sử dụng để đào tạo các hệ thống xử lý hình ảnh khác nhau. Khi các kỹ thuật học máy mới xuất hiện, MNIST vẫn là nguồn tài nguyên đáng tin cậy cho cả các nhà nghiên cứu và người học.\n\nCác hình ảnh đen trắng ban đầu đã được chuẩn hóa để vừa với hộp giới hạn 28x28 pixel và khử răng cưa, giới thiệu các mức thang độ xám. Cơ sở dữ liệu MNIST ban đầu chứa 60.000 hình ảnh đào tạo và 10.000 hình ảnh thử nghiệm, nhưng trong cuộc thi này, chúng tôi được cung cấp 42.000 hình ảnh đào tạo (70%) và 18.000 hình ảnh thử nghiệm (30%).\n\nCác kỹ thuật tiên tiến để tiếp cận bất kỳ vấn đề Phân loại hình ảnh nào thường dựa vào Mạng thần kinh chuyển đổi. Kiểu kiến ​​trúc này thực sự tốt trong việc trích xuất các mẫu từ hình ảnh. Tuy nhiên, vì MNIST là một bài toán đơn giản nên trong cuốn sổ tay này, chúng ta sẽ áp dụng các Thuật toán Học máy đơn giản (và chúng ta sẽ có được độ chính xác khá cao!).\n\nĐầu tiên, chúng tôi sẽ thực hiện Phân tích thành phần chính để giảm kích thước của dữ liệu (từ giá trị 784 pixel xuống còn 10 thành phần chính). Sau đó, chúng ta sẽ chạy thuật toán lười học yêu thích của mình: K-Nearest Neighbors. Combinatino này đạt được Độ chính xác thử nghiệm là 0,9271, một điều khá đáng chú ý !!!\n\nMục lục: 🔎 Đọc và khám phá dữ liệu 🔍:\n\nKiểm tra xem có thiếu giá trị nào không.\nKiểm tra phạm vi giá trị pixel.\nTrực quan hóa dữ liệu (Hình ảnh).\nKiểm tra xem mục tiêu có cân bằng không.\n🔶 PCA 🔶:\n\nGiới thiệu thuật toán PCA.\nTiền xử lý (quy mô) dữ liệu.\nTriển khai thuật toán PCA bằng thư viện Sklearn.\nNhúng các mô-đun vào Đường ống Sklearn.\nVẽ biểu đồ phương sai giải thích của các thành phần.\nHiển thị hai chiều của dữ liệu bằng cách sử dụng seaborn.\nTrực quan hóa dữ liệu ba chiều bằng cách sử dụng âm mưu.\nTrực quan hóa phép biến đổi nghịch đảo.\nKiểm tra xem các thành phần có tương quan với nhau không.\n🔶 Bộ phân loại KNN 🔶:\n\nXây dựng Bộ phân loại KNN bằng Sklearn.\nSử dụng kỹ thuật PCA làm công cụ trích xuất đặc trưng cho KNN.\nChạy Xác thực chéo tìm kiếm lưới để tối ưu hóa siêu tham số n_neighbors.\nSo sánh các số liệu huấn luyện và xác nhận để kiểm tra xem có trang bị quá mức không.\nVẽ ma trận nhầm lẫn.\nTính toán báo cáo phân loại.","metadata":{"_uuid":"85d40f0d-34d6-4e02-90f3-7ef7a60e5573","_cell_guid":"ddf0a964-5157-492a-9699-253bc2caec79","trusted":true}},{"cell_type":"code","source":"## IMPORT THE NECESSARY PYTHON LIBRARIES\n\n# Read & Explore Data\nimport numpy as np\nimport pandas as pd\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom sklearn.preprocessing import MinMaxScaler\n\n# Principal Component Analysis (PCA)\nfrom sklearn.decomposition import PCA\nfrom sklearn.pipeline import Pipeline\nimport plotly.express as px\n\n# K-Nearest Neighbor Classifier (KNN)\nfrom sklearn.neighbors import KNeighborsClassifier\nfrom sklearn.model_selection import GridSearchCV\nfrom sklearn.metrics import classification_report, \\\n confusion_matrix, ConfusionMatrixDisplay \n\n# VHT\nfrom scipy.ndimage import rotate","metadata":{"_uuid":"c382c9dc-2109-4ce6-b779-dc8c447da240","_cell_guid":"e169ec97-1f64-46cd-8099-1d09c757dccc","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:41:48.913838Z","iopub.execute_input":"2024-06-04T02:41:48.914752Z","iopub.status.idle":"2024-06-04T02:41:52.659289Z","shell.execute_reply.started":"2024-06-04T02:41:48.914706Z","shell.execute_reply":"2024-06-04T02:41:52.658009Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def func_array1d_rotate2d(Narray1d,rotate_angle,sizePixel):\n\n N = Narray1d.shape[0]\n\n Narray1d_rotate = np.zeros(Narray1d.shape)\n \n for i in range(N):\n image = np.reshape(Narray1d[i,:],(sizePixel,sizePixel))\n rotated_image = rotate(image,rotate_angle, reshape=False)\n rotated_array1d = np.reshape(rotated_image,(1,-1))\n Narray1d_rotate[i,:] = rotated_array1d\n \n return Narray1d_rotate","metadata":{"_uuid":"b362968d-9828-4257-92ec-b9e4e49a9cb0","_cell_guid":"8ace7936-3ee9-4fe9-b71e-8d7e3b7eee15","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:41:52.661298Z","iopub.execute_input":"2024-06-04T02:41:52.661840Z","iopub.status.idle":"2024-06-04T02:41:52.668988Z","shell.execute_reply.started":"2024-06-04T02:41:52.661808Z","shell.execute_reply":"2024-06-04T02:41:52.668138Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"----\n# Phần 2: Đọc dữ liệu và khai thác dữ liệu","metadata":{"_uuid":"a7abf7cf-fbfb-44f8-a5eb-77885d1283fb","_cell_guid":"1e5c33e8-406f-4da5-929d-034f87443736","trusted":true}},{"cell_type":"markdown","source":"Trong phần này, chúng tôi sẽ sử dụng thư viện Pandas Python nổi tiếng để đọc dữ liệu từ các tệp .csv.\n\nChúng ta sẽ trực quan hóa việc phân phối các giá trị pixel (biểu đồ) bằng thư viện Seaborn và Matplotlib Python. Chúng tôi cũng sẽ kiểm tra xem không có giá trị nào bị thiếu và các lớp mục tiêu có cân bằng hay không.\n\nCuối cùng, chúng ta sẽ trực quan hóa các hình ảnh bằng cách vẽ một hình ảnh từ mỗi lớp.","metadata":{"_uuid":"d4c82fac-2a10-4c14-a95a-653bd16872cd","_cell_guid":"deaebcb3-115c-4e6b-b843-4522382a052d","trusted":true}},{"cell_type":"code","source":"# Read the data files using Pandas\ndf_train = pd.read_csv('/kaggle/input/digit-recognizer/train.csv')\ndf_test = pd.read_csv('/kaggle/input/digit-recognizer/test.csv')","metadata":{"_uuid":"4da57b46-cc50-4ae9-9495-d0dbeb6be5ee","_cell_guid":"8967621a-0c55-49ff-b6e2-05d9efdeedf2","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:41:52.670635Z","iopub.execute_input":"2024-06-04T02:41:52.671132Z","iopub.status.idle":"2024-06-04T02:42:00.062872Z","shell.execute_reply.started":"2024-06-04T02:41:52.671072Z","shell.execute_reply":"2024-06-04T02:42:00.061797Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Check the size of the train and test DataFrames\nprint(\"df_train shape:\", df_train.shape) # 70%\nprint(\"df_test shape: \", df_test.shape) # 30%","metadata":{"_uuid":"5465a268-9ec4-4482-a0d7-e531d25f4917","_cell_guid":"da0e04b7-5ce3-48ec-be64-cee70c71af07","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:00.065973Z","iopub.execute_input":"2024-06-04T02:42:00.066449Z","iopub.status.idle":"2024-06-04T02:42:00.073200Z","shell.execute_reply.started":"2024-06-04T02:42:00.066407Z","shell.execute_reply":"2024-06-04T02:42:00.071957Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Fortunately there are no missing values in our data\nprint(\n \"Missing values in df_train: \", \\\n df_train.isna().sum().sum()\n)\nprint(\n \"Missing values in df_test: \", \\\n df_test.isna().sum().sum()\n)","metadata":{"_uuid":"f8324800-d007-4d21-9eb6-39bba50ce548","_cell_guid":"758f94a2-69c7-4c9b-b1a3-22cb0f0b1a43","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:00.074570Z","iopub.execute_input":"2024-06-04T02:42:00.074939Z","iopub.status.idle":"2024-06-04T02:42:00.134273Z","shell.execute_reply.started":"2024-06-04T02:42:00.074908Z","shell.execute_reply":"2024-06-04T02:42:00.132958Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's have a look at the first five training rows\ndf_train.head()","metadata":{"_uuid":"b1a7f2dc-6cec-4610-ba22-6ad5a7fb06ac","_cell_guid":"f1a8c252-31ba-46a1-b25c-c6ffff3fa2a9","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:00.135691Z","iopub.execute_input":"2024-06-04T02:42:00.136535Z","iopub.status.idle":"2024-06-04T02:42:00.168946Z","shell.execute_reply.started":"2024-06-04T02:42:00.136499Z","shell.execute_reply":"2024-06-04T02:42:00.167736Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# We can see that the first colum is the label,\n# and the rest of the columns are the 784 pixels\n# that represent 28x28 gray-scale images.\n# Let's separate X and Y\n\nX_train = df_train.drop(columns=['label']).values\ny_train = df_train['label'].values","metadata":{"_uuid":"48085649-c225-4e39-824f-82474ea004ff","_cell_guid":"939c1008-c85e-4ee0-bdab-bcccc4a252e4","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:00.173078Z","iopub.execute_input":"2024-06-04T02:42:00.173451Z","iopub.status.idle":"2024-06-04T02:42:00.308441Z","shell.execute_reply.started":"2024-06-04T02:42:00.173419Z","shell.execute_reply":"2024-06-04T02:42:00.307360Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's visualize the distribution of the pixel values\n\nax = sns.histplot(x=X_train.flatten(), color='#4895EE')\n_ = ax.set(title='Pixel Value Histogram', xlabel='Pixel Value')\n\n# As we can see, the 0 value represents the background\n# and the >0 values represent the digits in the images","metadata":{"_uuid":"f724afe4-d098-4f65-aaca-e97a096b8a32","_cell_guid":"36ec9986-72eb-4f04-abf8-84e2df3e3bfa","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:00.309895Z","iopub.execute_input":"2024-06-04T02:42:00.310383Z","iopub.status.idle":"2024-06-04T02:42:18.635163Z","shell.execute_reply.started":"2024-06-04T02:42:00.310334Z","shell.execute_reply":"2024-06-04T02:42:18.631687Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's check that the target is balanced.\n# It means that the ten categories in the target\n# have roughly the same representation in the data.\n\nax = sns.countplot(x=y_train, color='#4895EE')\nax.set(\n title='Target distribution is balanced? Yes!',\n xlabel='Label'\n)\n# The expected number of samples for a balanced \n# class distribution is 42000/10 = 4200\n_ = ax.axhline(y=4200, color='#12086F', linestyle='--')\n\n# Since the target classes are balances, we don't have\n# worry about sampling methods and biases in our models.","metadata":{"_uuid":"cd488e69-2d9d-4209-ac52-f11e13ab443b","_cell_guid":"d681064d-2dbe-4c2b-8075-9ce2c254d755","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:18.636823Z","iopub.execute_input":"2024-06-04T02:42:18.637192Z","iopub.status.idle":"2024-06-04T02:42:18.997266Z","shell.execute_reply.started":"2024-06-04T02:42:18.637162Z","shell.execute_reply":"2024-06-04T02:42:18.995923Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Next, let's visualize one image from each class.\n# We will use the matplotlib 'imshow' function.\n\nfig, axs = plt.subplots(nrows=2, ncols=5, figsize=(10,5))\n\nfor i in range(10):\n idx = np.where(y_train==i)[0][2]\n img = axs[int(i >= 5), i%5].imshow(\n X=X_train[idx].reshape(28, 28),\n cmap='Blues'\n )\n img.axes.get_yaxis().set_visible(False)\n img.axes.get_xaxis().set_visible(False)\n img.axes.set_title(f'Label: {i}')","metadata":{"_uuid":"9d6c909c-609f-4b5d-a907-2d75a3c17401","_cell_guid":"1fa8f38f-f989-4007-86c5-590612877157","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:19.002174Z","iopub.execute_input":"2024-06-04T02:42:19.002595Z","iopub.status.idle":"2024-06-04T02:42:19.767646Z","shell.execute_reply.started":"2024-06-04T02:42:19.002559Z","shell.execute_reply":"2024-06-04T02:42:19.766401Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"transform data","metadata":{"_uuid":"dd9d9d94-7ea8-4f8d-aa91-866959d703e3","_cell_guid":"48cf8b51-1a3e-4d0e-94e5-da53b3aa7c90","trusted":true}},{"cell_type":"code","source":"# -------------------------------------------------- declare\nnumRotate_prior = 10\n\nangleOffset_prior = -10\nangleRange_prior = 40\n\nX_train_pick2rotate = {\n 'index': np.array([3100, 3101, 3102, 3103, 3104, 3105, 3106, 3107, 3108, 3109, \n 3110, 3111, 3112, 3113, 3114, 3115, 3116, 3117, 3118, 3119,\n 3120, 3121, 3122, 3123, 3124, 3125, 3126, 3127, 3128, 3129,\n 3130, 3131, 3132, 3133, 3134, 3135, 3136, 3137, 3138, 3139,\n 3140, 3141, 3142, 3143, 3144, 3145, 3146, 3147, 3148, 3149,\n 3150, 3151, 3152, 3153, 3154, 3155, 3156, 3157, 3158, 3159,\n 3160, 3161, 3162, 3163, 3164, 3165, 3166, 3167, 3168, 3169,\n 3170, 3171, 3172, 3173, 3174, 3175, 3176, 3177, 3178, 3179,\n 3180, 3181, 3182, 3183, 3184, 3185, 3186, 3187, 3188, 3189,\n 3190, 3191, 3192, 3193, 3194, 3195, 3196, 3197, 3198, 3199]),\n 'offset': np.array([\n [-5, 30, -5, 20, 15, 10, 0, 10, 10, 20],\n [0, 30, 0, -5, 15, 5, -10, 0, 5, -5], \n [30, 5, 0, 0, 0, 5, 5, 0, 20, 0],\n [20, 0, 0, 0, 0, 10, 10, 0, 5, -5], \n [30, 10, -5, -10, 10, 10, -50, 0, -20, 10],\n [0, 10, 10, 0, 20, 0, -20, 10, 10, -10],\n [-10, 20, 10, 0, 15, 10, -30, 0, 10, 20],\n [20, 25, 0, 0, 0, -5, -5, 0, 5, -10], \n [0, 5, 0, 0, 10, 0, -20, -5, 0, 10],\n [0, 0, 0, -5, 0, -5, 0, 0, 10, 10],\n\n [-5, -10, -1, 5, 0, 10, 0, 10, 15, 0],\n [10, 30, 0, -5, 15, -5, -60, -10, 15, 10], \n [30, 5, 0, 20, 20, 5, -20, 0, -30, 0],\n [20, 0, 0, 0, 0, 10, -30, 0, 20, 20], \n [0, 30, -5, 0, 15, 15, -30, 0, 20, -10], \n [20, 20, 15, 10, 15, 0, -30, 0, 5, -5],\n [10, -20, -10, -10, 15, 10, -30, 5, 10, -25],\n [20, 10, 20, 0, 0, -5, -30, 20, 20, -10], \n [0, 20, 15, 10, 5, 0, 0, 0, 0, 0],\n [0, 40, -5, -5, -5, 5, -30, 0, 10, -10],\n\n [40, 25, 10, 10, 15, -5, 0, 5, 20, 10],\n [0, -20, -10, 0, 20, -5, -40, -5, 0, 15], \n [30, 10, -10, -20, 5, 5, -25, 20, -15, 0],\n [30, 0, 0, 5, 5, 10, -10, -5, -15, 0], \n [20, 5, -5, 10, 10, 15, 5, 0, 0, -10],\n [45, 0, -20, 0, 30, 10, -20, -10, 5, -10], \n [25, 10, 0, 0, 15, 15, -30, 0, 25, 15],\n [20, 5, 5, 5, 5, 0, 5, -10, 5, 10], \n [-85, 0, 0, 0, 10, -10, -30, -5, 30, 30],\n [20, 30, -15, -10, 20, 5, -40, 0, 5, 10],\n\n [30, -5, 0, -10, 15, 5, -20, 5, 15, 10],\n [15, 20, -10, -10, -20, 15, -20, -5, 5, 15], \n [-60, 30, -10, 30, 0, 15, -50, 20, -10, 5],\n [0, 15, 0, 20, 10, 15, -30, 10, 30, -5], \n [35, -5, -25, 5, 20, 10, -20, -10, 0, 10],\n [45, 0, 0, 10, -15, 25, 0, -10, 15, 0], \n [30, -10, -10, 10, 20, -10, -50, 5, -5, -10],\n [20, 10, 10, -25, 15, 10, -20, 0, 10, 15], \n [-85, 20, 0, 0, 0, 30, -50, 0, 20, 0],\n [40, 5, 0, 10, 20, 5, 20, 5, 5, -10],\n\n [10, 0, -5, 10, 10, 15, -40, 0, 0, 0],\n [30, 25, 0, 10, 20, 10, 0, 5, 20, 10], \n [40, 25, -5, 20, -10, 20, -50, 0, 10, 5],\n [0, 30, 0, -20, -10, 15, -50, 0, 20, 0], \n [40, 10, 0, 10, 30, 10, 10, 10, 10, -10],\n [30, 15, 0, 0, 0, 25, -20, 0, -5, 10], \n [35, 0, -20, 20, 10, 10, -50, -20, 0, 10],\n [20, 20, 0, 10, 15, 20, -20, 0, 25, 15], \n [10, 10, 10, 35, 25, 0, -30, -10, 0, 10],\n [40, 0, -15, -15, -5, 15, -30, 10, 10, 0],\n\n [80, 25, 20, 30, 0, 0, -30, -5, 5, 0],\n [45, 30, -5, -5, 15, 25, 0, 10, -10, 10], \n [35, 30, -10, 0, 10, 0, -10, 5, 20, -20],\n [30, -5, 0, 5, 0, 0, 0, 5, -10, -5], \n [25, 5, -10, 0, 5, 20, -20, 0, 30, 15],\n [25, 10, 10, -20, 15, 0, -20, 5, 20, 0],\n [0, 20, 0, -5, 15, 10, -50, 0, -10, 0],\n [30, 0, 5, 5, 0, -5, -60, -5, 10, -10], \n [20, 30, 10, 0, 10, 10, 10, -5, 5, 10],\n [20, 5, 0, 0, 10, -5, -15, 0, 15, 10],\n\n [10, 30, 0, -5, -10, 10, 20, -5, 5, 0],\n [40, 10, 10, -5, 30, 0, -20, 0, 0, 0], \n [0, -5, -10, -10, 30, -10, -10, 5, 0, -10],\n [30, -5, -10, 5, 10, 0, 10, -5, 25, 0], \n [30, 10, 10, -10, 5, -10, 15, 0, -5, 15],\n [60, 35, 5, 25, 10, 5, -15, 15, 10, 0],\n [0, 20, -5, 0, -10, 0, 0, -10, 25, 30],\n [35, 5, 5, 10, -10, 20, 20, -5, 25, 20], \n [30, 0, 0, -5, 20, 10, -30, -5, 5, 10],\n [0, 25, -5, 0, 5, -10, -15, 0-5, 20, 10],\n\n [-90, 0, 0, 0, -10, 15, 0, 5, 5, -20],\n [0, 10, 30, 20, -20, 20, -40, -5, 0, -5], \n [30, 15, 5, 5, 0, 10, 10, -5, 10, 10],\n [50, 15, -10, -5, 0, 20, 10, -20, 20, 10], \n [30, -5, -5, 0, -10, 10, -30, 0, 20, 15],\n [20, 10, 10, 0, 30, 10, -40, -10, -5, 10],\n [10, 25, 0, 10, 10, 0, -60, 5, -10, 25],\n [25, 0, 0, -30, 30, -5, -10, -10, 10, -10], \n [50, -10, 0, 20, 0, 20, -10, -10, -5, 5],\n [50, 30, 0, 5, 0, -5, -30, 0, 10, 20],\n\n [25, 0, 0, 20, 20, 15, 10, 0, 25, -5],\n [20, -10, 0, 20, 20, 20, -40, -5, 20, -5], \n [0, 0, 0, 5, 30, 0, 20, 5, 15, 5],\n [10, 0, 10, 5, 15, 10, 10, 0, 0, 10], \n [55, 5, 5, -5, 10, 10, -50, 0, 5, 15],\n [60, 0, 0, 0, 0, 10, -40, 0, 10, -10],\n [30, 10, 0, -10, 5, 20, -60, 5, -5, -20],\n [35, 20, 0, 25, 10, 5, 10, 20, 0, -10], \n [10, 15, 0, 15, 30, 20, -10, 10, 0, -10],\n [30, 5, -5, -5, -5, 20, -30, 0, -10, 20],\n\n [-40, 25, -5, 5, 30, 10, -5, -10, 0, 35],\n [30, 10, 0, 0, 45, -5, -40, -10, 35, 10],\n [25, 5, 0, 10, 15, 0, -45, -10, 20, 40],\n [25, -10, -25, -25, 10, 15, -30, -10, -5, 10], \n [20, 0, 0, -20, 10, 25, -70, 0, 0, 0],\n [25, 20, 0, 30, 35, 20, -25, 20, 10, 15], \n [-10, -20, 0, 30, -5, 20, -30, -10, 0, 10],\n [80, 35, -5, 0, -5, 10, -35, 0, 35, -25], \n [20, 10, 0, 0, 40, 20, -10, 0, 15, -5],\n [35, -10, 0, 0, -5, -5, 0, 0, 5, 0]\n ])\n}\n\nnumPickRotate = len(X_train_pick2rotate['index'])\n\nprint('numPickRotate = ', numPickRotate)\n\n# -------------------------------------------------- rotate\n\nX_sample_rotate = np.zeros((10,numRotate_prior,numPickRotate,X_train.shape[1]))\ny_sample_rotate = np.zeros((10,numRotate_prior,numPickRotate))\n\nfor k0 in range(10):\n for k1 in range(numRotate_prior):\n for k2 in range(numPickRotate):\n \n idxPick_X_train = X_train_pick2rotate['index'][k2]\n \n idx = np.where(y_train==k0)[0][idxPick_X_train]\n \n y_sample_rotate[k0,k1,k2] = k0 * numRotate_prior + k1\n \n angleRotate = (k1/numRotate_prior * angleRange_prior) + angleOffset_prior + X_train_pick2rotate['offset'][k2][k0]\n \n X_sample_rotate[k0,k1,k2,:] = func_array1d_rotate2d(X_train[idx][np.newaxis,:],angleRotate, 28)\n\n \nX_sample_rotate[X_sample_rotate < 10] = 0\n#X_sample_rotate[X_sample_rotate > 300] = 0\n\n# Let's normalize the pixel values to the interval [0,1]\n# Formula: MinMaxscale(X) = (X - X_min) / (X_max - X_min)\n\nprint(\"Before applying MinMax Scaler:\")\nprint(\" - Minimum value:\", X_sample_rotate.min())\nprint(\" - Maximum value:\", X_sample_rotate.max())\n\nminmax_scaler = MinMaxScaler()\n\nfor k0 in range(10):\n for k1 in range(numRotate_prior):\n X_sample_rotate[k0,k1,:,:] = minmax_scaler.fit_transform(X_sample_rotate[k0,k1,:,:]) * 255\n\nprint(\"\\nAfter applying MinMax Scaler:\")\nprint(\" - Minimum value:\", X_sample_rotate.min())\nprint(\" - Maximum value:\", X_sample_rotate.max())\n\n\n\n# Next, let's visualize one image from each class.\n# We will use the matplotlib 'imshow' function.\n\n# for k2 in range(numPickRotate):\nfor k2 in range(5,10):\n k1 = 0\n # k1 = numRotate_prior - 1\n\n fig, axs = plt.subplots(nrows=2, ncols=5, figsize=(10,5))\n\n for k0 in range(10):\n img = axs[int(k0 >= 5), k0%5].imshow(\n X=X_sample_rotate[k0,k1,k2,:].reshape(28, 28),\n cmap='Blues'\n )\n img.axes.get_yaxis().set_visible(False)\n img.axes.get_xaxis().set_visible(False)\n img.axes.set_title(f'Label: {k0}')\n \nX_train_rotate = np.reshape(X_sample_rotate, (-1,784))\ny_train_rotate = np.reshape(y_sample_rotate, (-1,1)).flatten()","metadata":{"_uuid":"2fd8cd6b-8db8-45dc-b5f7-6721c80f11fd","_cell_guid":"6b163e1f-7fee-43a6-b6f6-40c0d402a4fc","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:19.769468Z","iopub.execute_input":"2024-06-04T02:42:19.770238Z","iopub.status.idle":"2024-06-04T02:42:26.865442Z","shell.execute_reply.started":"2024-06-04T02:42:19.770203Z","shell.execute_reply":"2024-06-04T02:42:26.864143Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"---------\n# Phần 3: Phân tích thành phần chính","metadata":{"_uuid":"f67ae6c7-11f8-4b88-a23e-bf62fb9fdb43","_cell_guid":"6e9093b4-f586-4545-b99e-b59d492b7958","trusted":true}},{"cell_type":"markdown","source":"Phân tích thành phần chính (PCA) là một kỹ thuật thống kê để giảm tính chiều của tập dữ liệu. Điều này được thực hiện bằng cách chuyển đổi tuyến tính dữ liệu sang một hệ tọa độ mới trong đó (hầu hết) sự thay đổi trong dữ liệu có thể được mô tả với ít thứ nguyên hơn dữ liệu ban đầu.\n\nThành phần chính đầu tiên có thể được định nghĩa một cách tương đương là hướng tối đa hóa phương sai của dữ liệu dự kiến. Thành phần chính thứ i có thể được lấy làm hướng trực giao với thành phần chính thứ i-1 để tối đa hóa phương sai của dữ liệu dự kiến.\n\nsklearn.decomposition.PCA: Giảm kích thước tuyến tính bằng cách sử dụng Phân tách giá trị đơn lẻ của dữ liệu để chiếu dữ liệu đó vào không gian có chiều thấp hơn.\n\nỞ đây chúng tôi sẽ sử dụng 10 thành phần chính hàng đầu, giải thích gần 50% phương sai trong dữ liệu. Điều này có nghĩa là chúng tôi có thể biểu thị 50% phương sai của dữ liệu với không gian tính năng nhỏ hơn >98% so với dữ liệu ban đầu (10 tính năng so với 784 tính năng)","metadata":{"_uuid":"c9145741-6e7a-4419-bdb0-baca414ed4e7","_cell_guid":"54390ba7-75cb-4867-b577-97f23c85d6f2","trusted":true}},{"cell_type":"code","source":"# Let's normalize the pixel values to the interval [0,1]\n# Formula: MinMaxscale(X) = (X - X_min) / (X_max - X_min)\n\nprint(\"Before applying MinMax Scaler:\")\nprint(\" - Minimum value:\", X_train.min())\nprint(\" - Maximum value:\", X_train.max())\n\nminmax_scaler = MinMaxScaler()\nX_train_rotate_scaled = minmax_scaler.fit_transform(X_train_rotate)\n\nprint(\"\\nAfter applying MinMax Scaler:\")\nprint(\" - Minimum value:\", X_train_rotate_scaled.min())\nprint(\" - Maximum value:\", X_train_rotate_scaled.max())","metadata":{"_uuid":"aad53bf6-9b1f-4451-844d-ae51f509a68f","_cell_guid":"2f528db2-3b92-425b-a18c-2168dd8862d7","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:26.867035Z","iopub.execute_input":"2024-06-04T02:42:26.867449Z","iopub.status.idle":"2024-06-04T02:42:27.014475Z","shell.execute_reply.started":"2024-06-04T02:42:26.867416Z","shell.execute_reply":"2024-06-04T02:42:27.013079Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Actually, we will embed the MinMaxScaler and the PCA\n# modules into a Sklearn Pipeline.\n\npca = Pipeline(steps=[\n ('scaler', MinMaxScaler()),\n ('pca', PCA(n_components=40))\n])\n\nX_train_rotate_pca = pca.fit_transform(X=X_train_rotate)\n\nprint('X_train_pca shape:', X_train_rotate_pca.shape)","metadata":{"_uuid":"6cae5170-3af3-4ed9-a9d8-372ef224e9ba","_cell_guid":"e423b865-ef1d-4c48-9152-917f8b258060","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:27.016022Z","iopub.execute_input":"2024-06-04T02:42:27.016406Z","iopub.status.idle":"2024-06-04T02:42:28.666905Z","shell.execute_reply.started":"2024-06-04T02:42:27.016366Z","shell.execute_reply":"2024-06-04T02:42:28.665387Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's extract the explained variance ratio\n# And let's compute their cumulative sum\n\npca_object = pca.named_steps['pca']\nvar_ratio = pd.Series(\n data=pca_object.explained_variance_ratio_,\n index=range(1, 41)\n)\n\ncumsum_var_ratio = var_ratio.cumsum()","metadata":{"_uuid":"68b1ffab-f2de-43fe-a655-e04c6060ec60","_cell_guid":"c3cdf97b-6136-45e0-94c6-bdf23e750bf0","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:28.669380Z","iopub.execute_input":"2024-06-04T02:42:28.670553Z","iopub.status.idle":"2024-06-04T02:42:28.689600Z","shell.execute_reply.started":"2024-06-04T02:42:28.669885Z","shell.execute_reply":"2024-06-04T02:42:28.687663Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Plot the explained variance ratio for each component (left)\n# And plot their cumulative sum (right)\n\nfig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15,5))\n\nsns.barplot(\n x=var_ratio.index, \n y=var_ratio.values, \n color='#4895EE', \n ax=ax1\n)\nax1.set(\n title='Explained Variance Ratio PCA',\n xlabel='Component',\n ylabel='Var Ratio',\n)\nax1.set_ylim(0, 0.5)\n\nsns.barplot(\n x=cumsum_var_ratio.index, \n y=cumsum_var_ratio.values, \n color='#4895EE', \n ax=ax2\n)\nax2.set(\n title='Cumulative Explained Variance Ratio PCA',\n xlabel='Component',\n ylabel='Cumulative Var Ratio',\n)\n_ = ax2.set_ylim(0, 0.5)","metadata":{"_uuid":"9ea36679-af0c-40f1-bda2-641ef9c73a4e","_cell_guid":"6fe44c61-1ac9-4157-968d-022d42184ff6","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:28.692892Z","iopub.execute_input":"2024-06-04T02:42:28.694065Z","iopub.status.idle":"2024-06-04T02:42:30.171718Z","shell.execute_reply.started":"2024-06-04T02:42:28.694002Z","shell.execute_reply":"2024-06-04T02:42:30.170323Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Let's put the components in Pandas DataFrame format\n# This way we can use the seaborn and plotly libraries\n# to visualize the data.\n# (these packages work better with Pandas DataFrames)\n\ndf_train_pca = pd.DataFrame(\n data=X_train_rotate_pca[:,:3], \n columns=['x', 'y', 'z']\n)\n# Let's also add the target to set the colors\ndf_train_pca['target'] = y_train_rotate\n\nprint(df_train_pca.shape)\ndf_train_pca.head()","metadata":{"_uuid":"9796e066-c6af-49c5-af8f-8ebde5e07430","_cell_guid":"4fb22220-cc87-410b-a637-55bc57791987","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:30.173654Z","iopub.execute_input":"2024-06-04T02:42:30.174460Z","iopub.status.idle":"2024-06-04T02:42:30.190766Z","shell.execute_reply.started":"2024-06-04T02:42:30.174424Z","shell.execute_reply":"2024-06-04T02:42:30.189963Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Two-dimensional visualization of the data\n# Here the Top2 principal components are plotted\n\nsns.set_style(\"dark\")\nax = sns.scatterplot(\n data=df_train_pca,\n x='x', \n y='y', \n hue='target', \n palette='Blues',\n)\nax.set(\n title='PCA Two-Dimensional Projection',\n xlabel='1st Comp.',\n ylabel='2nd Comp.'\n)\nax.axhline(y=0, linestyle='--', color='#FFBF00', zorder=-1)\nax.axvline(x=0, linestyle='--', color='#FFBF00', zorder=-1)","metadata":{"_uuid":"30135dee-d7d0-4dc8-88f7-3a88a135ad5a","_cell_guid":"147fd5ee-e632-4611-b6de-35d64c6946e9","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:30.192472Z","iopub.execute_input":"2024-06-04T02:42:30.193052Z","iopub.status.idle":"2024-06-04T02:42:31.727063Z","shell.execute_reply.started":"2024-06-04T02:42:30.193019Z","shell.execute_reply":"2024-06-04T02:42:31.726187Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Now, let's apply the inverse_transform to compute\n# the approximation of images based on the 10 principal\n# compontents.\n# We can see that with just 10 components we convey a large\n# part of the information of each image.\n\npca_object = pca.named_steps['pca']\n\nfig, axs = plt.subplots(nrows=10, ncols=2, figsize=(6,20))\n\nfor i in range(10):\n idx = np.where(y_train==i)[0][0]\n original_img = axs[i, 0].imshow(\n X=X_train_rotate[idx].reshape(28, 28),\n cmap='Blues'\n )\n original_img.axes.get_yaxis().set_visible(False)\n original_img.axes.get_xaxis().set_visible(False)\n original_img.axes.set_title(f'Original [{i}]:')\n \n approx_img_ = pca_object.inverse_transform(X=X_train_rotate_pca[idx])\n approx_img = axs[i, 1].imshow(\n X=approx_img_.reshape(28, 28),\n cmap='Blues'\n )\n approx_img.axes.get_yaxis().set_visible(False)\n approx_img.axes.get_xaxis().set_visible(False)\n approx_img.axes.set_title(f'Approximation [{i}]:')","metadata":{"_uuid":"77146579-275e-4359-8ec9-7d175da30220","_cell_guid":"329c95b3-2406-402c-b1f8-38487c4fa2f3","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:31.728461Z","iopub.execute_input":"2024-06-04T02:42:31.729447Z","iopub.status.idle":"2024-06-04T02:42:32.825489Z","shell.execute_reply.started":"2024-06-04T02:42:31.729413Z","shell.execute_reply":"2024-06-04T02:42:32.824339Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Show that the PCA components are uncorrelated\n# (This statement is true by definition and construction\n# of the Principal Component Analysis algorithm)\n\ncorr = np.corrcoef(x=X_train_rotate_pca, rowvar=False)\ncorr = np.around(corr, decimals=2)\n\nax = sns.heatmap(\n data=corr, \n vmax=1, \n vmin=-1,\n square=True,\n linewidths=.5,\n annot=True,\n cmap='Blues',\n)\n\n_ = ax.set(\n title='Correlation PCA Components [Uncorrelated]',\n xlabel='Components',\n ylabel='Components'\n)","metadata":{"_uuid":"8c3fa8bb-61ab-43dc-bc7e-e5fecc13e6e2","_cell_guid":"ab321270-3628-423b-8fd6-a7ebaf6ccfe5","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:32.826882Z","iopub.execute_input":"2024-06-04T02:42:32.827742Z","iopub.status.idle":"2024-06-04T02:42:37.518272Z","shell.execute_reply.started":"2024-06-04T02:42:32.827709Z","shell.execute_reply":"2024-06-04T02:42:37.517087Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"----\n# Phần 4: Sử dụng K-Nearest Neighbors Classifier phân loại 10 thành phần chính hàng đầu từ PCA.","metadata":{"_uuid":"3d69b1a4-faf3-4405-b740-9d5fc82b8424","_cell_guid":"16cefed3-22ad-4c84-ad87-230ed47df34d","trusted":true}},{"cell_type":"markdown","source":"Thuật toán K-Nearest Neighbors là một thuật toán phân loại học tập có giám sát, không tham số, sử dụng độ gần để phân loại hoặc dự đoán về việc nhóm một điểm dữ liệu riêng lẻ. Dự đoán: Nhãn lớp được chỉ định trên cơ sở đa số phiếu bầu. nhãn được thể hiện thường xuyên nhất xung quanh một điểm dữ liệu nhất định.\n\nTrong dự án này, đầu vào của Bộ phân loại KNN sẽ là 10 thành phần chính hàng đầu từ PCA.\n\nsklearn.neighbors.KNeighborsClassifier: Trình phân loại thực hiện bỏ phiếu k-hàng xóm gần nhất. Chúng tôi sẽ chạy GridSearchCV để tối ưu hóa siêu tham số n_neighbors.","metadata":{"_uuid":"1abd82c5-f4e1-4142-acb2-bea45e42488e","_cell_guid":"b9e0603d-4dc8-4761-baad-a34cf6709975","trusted":true}},{"cell_type":"markdown","source":"mô hình học máy trên tập dữ liệu Train","metadata":{"_uuid":"334a6d45-a6b2-499e-a95b-2477800b968f","_cell_guid":"e1b5a3d3-ced9-409c-b6f9-b07e7143bc66","trusted":true}},{"cell_type":"code","source":"knn_clf = Pipeline(steps=[\n ('scaler', MinMaxScaler()),\n ('pca', PCA(n_components=10)),\n ('knn', KNeighborsClassifier())\n])\n\nknn_clf_params = {\n 'knn__n_neighbors': [2, 4, 8, 16, 32]\n}\n\nknn_clf_gs = GridSearchCV(\n estimator=knn_clf, \n param_grid=knn_clf_params,\n scoring='accuracy',\n n_jobs=-1,\n cv=5,\n verbose=1,\n return_train_score=True\n)\n\nknn_clf_gs.fit(X=X_train_rotate_pca, y=y_train_rotate)\n\nprint('best PCA+KNN params:', knn_clf_gs.best_params_)\nprint('best PCA+KNN cv accuracy:', knn_clf_gs.best_score_)","metadata":{"_uuid":"7154a244-1406-4e55-9062-2a876ffcfb42","_cell_guid":"f8b23485-498b-446c-83a4-fdf1e50b8285","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:37.519706Z","iopub.execute_input":"2024-06-04T02:42:37.520068Z","iopub.status.idle":"2024-06-04T02:42:56.266639Z","shell.execute_reply.started":"2024-06-04T02:42:37.520037Z","shell.execute_reply":"2024-06-04T02:42:56.264644Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Plot the Training and Validation Accuracies\n# for each Hyperparamter value tested during the\n# Grid Search.\n# This plot allows us to check for overfitting.\n\nfig, ax = plt.subplots()\n\nax.plot(\n range(len(knn_clf_params['knn__n_neighbors'])), \n knn_clf_gs.cv_results_['mean_train_score'],\n label='train',\n color='blue'\n)\nax.plot(\n range(len(knn_clf_params['knn__n_neighbors'])),\n knn_clf_gs.cv_results_['mean_test_score'],\n label='test',\n color='orange'\n)\nax.set_xticks(\n ticks=range(len(knn_clf_params['knn__n_neighbors'])),\n labels=knn_clf_params['knn__n_neighbors']\n)\n\nii = knn_clf_params['knn__n_neighbors'].index(\n knn_clf_gs.best_params_['knn__n_neighbors']\n)\n\nplt.axvline(x=ii, linestyle='--', color='red', label='best', alpha=0.3)\nplt.legend()\nplt.xlabel(\"'n_neighbors' vals\")\nplt.ylabel('Accuracy')\nplt.title(\"PCA+KNN GridSearchCV Results\")\nplt.show()","metadata":{"_uuid":"43e1d701-a89a-463f-b860-3afbf14bbf36","_cell_guid":"b1316942-7e1c-4f06-8742-d3eb8f7918c3","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:56.269370Z","iopub.execute_input":"2024-06-04T02:42:56.270490Z","iopub.status.idle":"2024-06-04T02:42:56.850986Z","shell.execute_reply.started":"2024-06-04T02:42:56.270421Z","shell.execute_reply":"2024-06-04T02:42:56.849648Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"---\n# Phần 5: Tạo một tệp tin chứa các dự đoán của mô hình học máy trên tập dữ liệu Test","metadata":{"_uuid":"b609f283-8a5c-477c-9c29-941acfedc9aa","_cell_guid":"5d9eb764-a90a-48b5-b962-be1b347e4925","trusted":true}},{"cell_type":"code","source":"from sklearn.metrics import accuracy_score\n\n# Compute and display the Confusion Matrix\n# Note: the entire training dataset is used here\n\nX_val = X_train_rotate_pca\nY_val = y_train_rotate\n\nbest_cv_clf = knn_clf_gs.best_estimator_\ny_pred = best_cv_clf.predict(X=X_val)\n\ny_val_number = Y_val // numRotate_prior\ny_pred_number = y_pred // numRotate_prior\n\ny_val_angle = (Y_val % numRotate_prior)\ny_pred_angle = (y_pred % numRotate_prior) \n\nprint('KNN number + angle, Acc({:.2f}%)', accuracy_score(Y_val, y_pred))\nprint('KNN number, Acc({:.2f}%)', accuracy_score(y_val_number, y_pred_number))\nprint('KNN angle, Acc({:.2f}%)', accuracy_score(y_val_angle, y_pred_angle))\n\n# print('y_val_number =', y_val_number)\n# print('y_pred_number =', y_pred_number)\n\ny_val_angle_degree = y_val_angle/numRotate_prior * angleRange_prior + angleOffset_prior\ny_pred_angle_degree = y_pred_angle/numRotate_prior * angleRange_prior + angleOffset_prior\n\n# print('y_val_angle_degree =', y_val_angle_degree)\n# print('y_pred_angle_degree =', y_pred_angle_degree)\n\nconf_matrix = confusion_matrix(y_val_number, y_pred_number)\nconf_matrix_disp = ConfusionMatrixDisplay(conf_matrix)\nconf_matrix_disp.plot(cmap='Blues')\n\nconf_matrix = confusion_matrix(y_val_angle, y_pred_angle)\nconf_matrix_disp = ConfusionMatrixDisplay(conf_matrix)\nconf_matrix_disp.plot(cmap='Blues')","metadata":{"_uuid":"62d7493a-d5c0-42ca-ad56-df419c3367ee","_cell_guid":"fdfdc4ce-7aa6-4c8a-b7c9-8a069cb98054","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:42:56.852690Z","iopub.execute_input":"2024-06-04T02:42:56.853610Z","iopub.status.idle":"2024-06-04T02:43:00.708883Z","shell.execute_reply.started":"2024-06-04T02:42:56.853567Z","shell.execute_reply":"2024-06-04T02:43:00.707610Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_val = pca.transform(X_train)\nY_val = y_train\n\nbest_cv_clf = knn_clf_gs.best_estimator_\ny_pred = best_cv_clf.predict(X=X_val)\n\ny_pred_number = y_pred\ny_val_number = Y_val\n\nprint('KNN number, Acc({:.2f}%)', accuracy_score(y_val_number, y_pred_number))\n\nconf_matrix = confusion_matrix(y_val_number, y_pred_number)\nconf_matrix_disp = ConfusionMatrixDisplay(conf_matrix)\nconf_matrix_disp.plot(cmap='Blues')\n\ny_pred_angle = (y_pred % numRotate_prior)\n\ny_pred_angle_degree = y_pred_angle/numRotate_prior * angleRange_prior + angleOffset_prior\n\nprint('y_pred_angle_degree =', y_pred_angle_degree)","metadata":{"_uuid":"da419e38-d545-49ee-a8b6-32f3fa9760c3","_cell_guid":"fa5652c2-d596-4880-8396-b6e77177f4ff","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:00.710719Z","iopub.execute_input":"2024-06-04T02:43:00.711208Z","iopub.status.idle":"2024-06-04T02:43:35.928535Z","shell.execute_reply.started":"2024-06-04T02:43:00.711167Z","shell.execute_reply":"2024-06-04T02:43:35.927342Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Now, let's apply the inverse_transform to compute\n# the approximation of images based on the 10 principal\n# compontents.\n# We can see that with just 10 components we convey a large\n# part of the information of each image.\n\npca_object = pca.named_steps['pca']\n\nfig, axs = plt.subplots(nrows=10, ncols=2, figsize=(6,20))\n\nfor i in range(10):\n idx = np.where(y_train == i )[0][0]\n original_img = axs[i, 0].imshow(\n X=X_train[idx].reshape(28, 28),\n cmap='Blues'\n )\n original_img.axes.get_yaxis().set_visible(False)\n original_img.axes.get_xaxis().set_visible(False)\n original_img.axes.set_title(f'Original [{i}]:')\n \n approx_img_ = pca_object.inverse_transform(X=X_val[idx])\n approx_img = axs[i, 1].imshow(\n X=approx_img_.reshape(28, 28),\n cmap='Blues'\n )\n approx_img.axes.get_yaxis().set_visible(False)\n approx_img.axes.get_xaxis().set_visible(False)\n approx_img.axes.set_title(f'Predicted angle [{y_pred_angle_degree[idx]}]:')","metadata":{"_uuid":"bc4e4c18-00b8-4441-900b-83786685d798","_cell_guid":"1393d26a-3a87-4a77-a642-c5b68c94fec3","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:35.930225Z","iopub.execute_input":"2024-06-04T02:43:35.930586Z","iopub.status.idle":"2024-06-04T02:43:37.015252Z","shell.execute_reply.started":"2024-06-04T02:43:35.930554Z","shell.execute_reply":"2024-06-04T02:43:37.014291Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n
\n

\nLight CNN Classifier\n

\n
\n
","metadata":{"_uuid":"a1446365-443d-48e8-aa24-514805f2df61","_cell_guid":"f6dca012-b643-484b-8f63-b0e218c6f1e2","trusted":true}},{"cell_type":"markdown","source":"","metadata":{"_uuid":"5a238d9f-7fbe-4d2a-85c7-b38f1793fe47","_cell_guid":"d606a8da-01ba-4650-b234-842037898d97","trusted":true}},{"cell_type":"markdown","source":"CNN là một cách rất hiệu quả cho các nhiệm vụ thị giác máy tính. \nMạng nơ-ron tích chập (CNN, hoặc ConvNet) là một lớp Mạng nơ-ron nhân tạo (ANN), được áp dụng phổ biến nhất để phân tích hình ảnh trực quan. Chúng còn được gọi là Shift Invariant hoặc Space Invariant Artificial Neural Networks (SIANN), dựa trên kiến trúc trọng lượng chia sẻ của các hạt nhân hoặc bộ lọc tích chập trượt dọc theo các tính năng đầu vào và cung cấp các phản hồi đẳng biến dịch được gọi là bản đồ tính năng. Phản trực giác, hầu hết mạng lưới thần kinh tích chập chỉ là đẳng biến, trái ngược với bất biến, để dịch. Chúng có các ứng dụng trong nhận dạng hình ảnh và video, hệ thống đề xuất, phân loại hình ảnh, phân đoạn hình ảnh, phân tích hình ảnh y tế, xử lý ngôn ngữ tự nhiên, giao diện não-máy tính và chuỗi thời gian tài chính.","metadata":{"_uuid":"3f62053b-0364-4ff6-bb7b-5deba2a02280","_cell_guid":"1c349770-f9d8-4f7f-8efb-d87b402f2f10","trusted":true}},{"cell_type":"code","source":"import plotly.express as px\nfrom plotly.offline import init_notebook_mode, iplot\nimport plotly.graph_objects as go\nimport warnings\nwarnings.filterwarnings('ignore')","metadata":{"execution":{"iopub.status.busy":"2024-06-04T02:43:37.016471Z","iopub.execute_input":"2024-06-04T02:43:37.016787Z","iopub.status.idle":"2024-06-04T02:43:37.046709Z","shell.execute_reply.started":"2024-06-04T02:43:37.016759Z","shell.execute_reply":"2024-06-04T02:43:37.045708Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def plot_confusion_matrix(cm, labels, title):\n # cm : confusion matrix list(list)\n # labels : name of the data list(str)\n # title : title for the heatmap\n data = go.Heatmap(z=cm, y=labels, x=labels)\n annotations = []\n for i, row in enumerate(cm):\n for j, value in enumerate(row):\n annotations.append(\n {\n \"x\": labels[i],\n \"y\": labels[j],\n \"font\": {\"color\": \"white\"},\n \"text\": str(value),\n \"xref\": \"x1\",\n \"yref\": \"y1\",\n \"showarrow\": False\n }\n )\n layout = {\n \"title\": title,\n \"xaxis\": {\"title\": \"Predicted value\"},\n \"yaxis\": {\"title\": \"Real value\"},\n \"annotations\": annotations\n }\n fig = go.Figure(data=data, layout=layout)\n return fig","metadata":{"execution":{"iopub.status.busy":"2024-06-04T02:43:37.048031Z","iopub.execute_input":"2024-06-04T02:43:37.048589Z","iopub.status.idle":"2024-06-04T02:43:37.057083Z","shell.execute_reply.started":"2024-06-04T02:43:37.048559Z","shell.execute_reply":"2024-06-04T02:43:37.055811Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# from tensorflow.keras.utils import np_utils\nfrom tensorflow.keras import utils\nfrom tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping\nfrom tensorflow.keras import datasets, layers, models","metadata":{"_uuid":"17fc4707-d986-4950-b0fe-ac152932ac00","_cell_guid":"0d0efe69-4e03-4204-86d8-3ee2e5125830","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:37.058525Z","iopub.execute_input":"2024-06-04T02:43:37.058932Z","iopub.status.idle":"2024-06-04T02:43:51.420749Z","shell.execute_reply.started":"2024-06-04T02:43:37.058903Z","shell.execute_reply":"2024-06-04T02:43:51.419640Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_train_cnn = X_train_rotate.reshape(-1,28,28,1)\n# X_val = np.array(pca.transform(X_train)).reshape(-1,28,28,1)\ny_train_cnn = utils.to_categorical(np.array(y_train_rotate// numRotate_prior))\nX_val = X_train.reshape(-1,28,28,1)\nY_val = utils.to_categorical(np.array(y_train))","metadata":{"_uuid":"598e24f0-4665-4486-ba5d-fb550022bc96","_cell_guid":"b20ca6f3-b288-42fe-8210-9e804c27268e","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:51.427545Z","iopub.execute_input":"2024-06-04T02:43:51.428304Z","iopub.status.idle":"2024-06-04T02:43:51.438919Z","shell.execute_reply.started":"2024-06-04T02:43:51.428266Z","shell.execute_reply":"2024-06-04T02:43:51.437639Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_train_cnn.shape","metadata":{"execution":{"iopub.status.busy":"2024-06-04T02:43:51.440297Z","iopub.execute_input":"2024-06-04T02:43:51.440653Z","iopub.status.idle":"2024-06-04T02:43:51.483800Z","shell.execute_reply.started":"2024-06-04T02:43:51.440623Z","shell.execute_reply":"2024-06-04T02:43:51.482927Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# y_train_rotate// numRotate_prior","metadata":{"execution":{"iopub.status.busy":"2024-06-04T02:43:51.485499Z","iopub.execute_input":"2024-06-04T02:43:51.485861Z","iopub.status.idle":"2024-06-04T02:43:51.495600Z","shell.execute_reply.started":"2024-06-04T02:43:51.485831Z","shell.execute_reply":"2024-06-04T02:43:51.494342Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# X_train_rotate = np.array(X_train_rotate).reshape(-1,28,28,1)\n# # X_val = np.array(pca.transform(X_train)).reshape(-1,28,28,1)\n# X_val = np.array(X_test_crop).reshape(-1,28,28,1)\n# y_train_rotate = utils.to_categorical(np.array(y_train_rotate))\n# Y_val = utils.to_categorical(np.array(y_test_crop))","metadata":{"_uuid":"f978abbd-867f-46a5-b6ec-35f64e740f64","_cell_guid":"9d27fd4b-f9a8-4c46-a856-7e8970246998","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:51.497310Z","iopub.execute_input":"2024-06-04T02:43:51.497646Z","iopub.status.idle":"2024-06-04T02:43:51.509881Z","shell.execute_reply.started":"2024-06-04T02:43:51.497614Z","shell.execute_reply":"2024-06-04T02:43:51.508898Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = models.Sequential()\nmodel.add(layers.Conv2D(filters = 32, kernel_size = (5,5),padding = 'Same', \n activation ='relu', input_shape = (28,28,1)))\nmodel.add(layers.Conv2D(filters = 32, kernel_size = (5,5),padding = 'Same', \n activation ='relu'))\nmodel.add(layers.MaxPool2D(pool_size=(2,2)))\nmodel.add(layers.BatchNormalization())\nmodel.add(layers.Dropout(0.25))\nmodel.add(layers.Conv2D(filters = 64, kernel_size = (3,3),padding = 'Same', \n activation ='relu'))\nmodel.add(layers.Conv2D(filters = 64, kernel_size = (3,3),padding = 'Same', \n activation ='relu'))\nmodel.add(layers.MaxPool2D(pool_size=(2,2), strides=(2,2)))\nmodel.add(layers.BatchNormalization())\nmodel.add(layers.Dropout(0.25))\nmodel.add(layers.Flatten())\nmodel.add(layers.Dense(256, activation = \"relu\"))\nmodel.add(layers.Dropout(0.5))\nmodel.add(layers.BatchNormalization())\nmodel.add(layers.Dense(10, activation = \"softmax\"))\n\nmodel.summary()","metadata":{"_uuid":"d30da279-3261-4e97-9b79-6e2a01d703d4","_cell_guid":"43ead369-05c5-4e2d-960c-d5773b8a7f7d","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:51.511284Z","iopub.execute_input":"2024-06-04T02:43:51.512200Z","iopub.status.idle":"2024-06-04T02:43:51.906058Z","shell.execute_reply.started":"2024-06-04T02:43:51.512155Z","shell.execute_reply":"2024-06-04T02:43:51.904831Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model_path = './best_model.h5'\ncheckpointer = ModelCheckpoint(\n filepath=model_path,\n monitor='val_accuracy',\n verbose=1,\n mode='max',\n save_best_only=True)\n\nmodel.compile(loss='categorical_crossentropy',\n optimizer='adam',\n metrics=['accuracy'])","metadata":{"_uuid":"d131aa49-918d-43dd-be19-62d3110ca17a","_cell_guid":"db889d42-b8bb-4667-8f07-b831a3fbcf38","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:51.907939Z","iopub.execute_input":"2024-06-04T02:43:51.909099Z","iopub.status.idle":"2024-06-04T02:43:51.936581Z","shell.execute_reply.started":"2024-06-04T02:43:51.909063Z","shell.execute_reply":"2024-06-04T02:43:51.935197Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"history = model.fit(X_train_cnn, y_train_cnn, validation_data = (X_val, Y_val), epochs=100, batch_size = 128, callbacks=[checkpointer])","metadata":{"_uuid":"620687bc-89d9-40dc-b133-006fef4e7c78","_cell_guid":"cd8420f8-3c35-407f-93de-b1691d1b534f","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T02:43:51.938471Z","iopub.execute_input":"2024-06-04T02:43:51.938856Z","iopub.status.idle":"2024-06-04T04:08:16.330595Z","shell.execute_reply.started":"2024-06-04T02:43:51.938811Z","shell.execute_reply":"2024-06-04T04:08:16.329576Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import plotly.graph_objects as go\nfrom plotly.subplots import make_subplots\nimport numpy as np\n\nfig = go.Figure()\n\nlabels = ['Loss', 'Accuracy']\ncolors = ['rgb(67,67,67)', 'rgb(49,130,189)']\n\nmode_size = [8, 8]\nline_size = [2, 2]\n\nx_data = np.vstack((np.arange(1, 51),)*2)\n\ny_data = np.array([\n np.array(history.history['loss']), \n np.array(history.history['accuracy'])\n])\n\nfor i in range(0, 2):\n fig.add_trace(go.Scatter(x=x_data[i], y=y_data[i], mode='lines',\n name=labels[i],\n line=dict(color=colors[i], width=line_size[i]),\n connectgaps=True,\n ))\n\n # endpoints\n fig.add_trace(go.Scatter(\n x=[x_data[i][0], x_data[i][-1]],\n y=[y_data[i][0], y_data[i][-1]],\n mode='markers',\n marker=dict(color=colors[i], size=mode_size[i])\n ))\n\nfig.update_layout(\n xaxis=dict(\n showline=True,\n showgrid=False,\n showticklabels=True,\n linecolor='rgb(204, 204, 204)',\n linewidth=2,\n ticks='outside',\n tickfont=dict(\n family='Arial',\n size=12,\n color='rgb(82, 82, 82)',\n ),\n ),\n yaxis=dict(\n showgrid=False,\n zeroline=False,\n showline=False,\n showticklabels=False,\n ),\n autosize=False,\n margin=dict(\n autoexpand=False,\n l=100,\n r=20,\n t=110,\n ),\n showlegend=False,\n plot_bgcolor='white'\n)\n\nannotations = []\n\n# Adding labels\nfor y_trace, label, color in zip(y_data, labels, colors):\n # labeling the left_side of the plot\n annotations.append(dict(xref='paper', x=0.05, y=y_trace[0],\n xanchor='right', yanchor='middle',\n text=label + ' {:.2f}'.format(y_trace[0]),\n font=dict(family='Arial',\n size=16),\n showarrow=False))\n # labeling the right_side of the plot\n annotations.append(dict(xref='paper', x=0.95, y=y_trace[49],\n xanchor='left', yanchor='middle',\n text='{:.4f}%'.format(y_trace[49]),\n font=dict(family='Arial',\n size=16),\n showarrow=False))\n# Title\nannotations.append(dict(xref='paper', yref='paper', x=0.0, y=1.05,\n xanchor='left', yanchor='bottom',\n text='Training Loss & Accuracy',\n font=dict(family='Arial',\n size=30,\n color='rgb(37,37,37)'),\n showarrow=False))\n# Source\nannotations.append(dict(xref='paper', yref='paper', x=0.5, y=-0.1,\n xanchor='center', yanchor='top',\n text='Epochs',\n font=dict(family='Arial',\n size=12,\n color='rgb(150,150,150)'),\n showarrow=False))\n\nfig.update_layout(annotations=annotations,height=600, width=1000)\n\nfig.show()","metadata":{"_uuid":"b2a98a55-6032-40d1-8bbf-bf6b35840730","_cell_guid":"7fcbf806-1db3-45b9-8bc0-3abf08dfabd9","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T04:08:16.332603Z","iopub.execute_input":"2024-06-04T04:08:16.332996Z","iopub.status.idle":"2024-06-04T04:08:16.878676Z","shell.execute_reply.started":"2024-06-04T04:08:16.332963Z","shell.execute_reply":"2024-06-04T04:08:16.877256Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"_uuid":"c51476fa-c9d6-4cbc-a5ad-f6ff9706d70e","_cell_guid":"8bdfa1eb-cd3b-4e48-a377-02b2b66506e1","collapsed":false,"jupyter":{"outputs_hidden":false},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from keras.models import load_model\nmodel = load_model('./best_model.h5')\ny_pred = model.predict(X_val)\nM = confusion_matrix(np.argmax(Y_val,1), np.argmax(y_pred,1))\nplot_confusion_matrix(M, sorted(np.unique(np.argmax(Y_val,1))), 'Simple CNN Confusion Matrix, Acc({:.2f}%)'.format(accuracy_score(np.argmax(Y_val,1), np.argmax(y_pred,1))*100))","metadata":{"_uuid":"57a8380a-11c2-45fb-8190-4e9da5bce1ed","_cell_guid":"bdeca8dc-2a69-4598-b508-873212a964dd","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T04:08:16.880837Z","iopub.execute_input":"2024-06-04T04:08:16.881335Z","iopub.status.idle":"2024-06-04T04:08:56.622093Z","shell.execute_reply.started":"2024-06-04T04:08:16.881289Z","shell.execute_reply":"2024-06-04T04:08:56.620814Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"
\n
\n

\nChecking Failure Cases\n

\n
\n
","metadata":{"_uuid":"8f029ba0-23b9-4fa0-83a3-eae4a5e4b39b","_cell_guid":"d955f566-e62b-4f51-a801-3de790bca12a","trusted":true}},{"cell_type":"code","source":"n = np.where(np.argmax(y_pred,1)!=np.argmax(Y_val,1))\nn = np.array(n).reshape(-1)\nn","metadata":{"_uuid":"042a5e34-2598-420b-bb05-79a34d48eac3","_cell_guid":"3cea1159-64d0-4e09-88e9-7b2f3a4fc1fb","collapsed":false,"jupyter":{"outputs_hidden":false},"execution":{"iopub.status.busy":"2024-06-04T04:08:56.624037Z","iopub.execute_input":"2024-06-04T04:08:56.624743Z","iopub.status.idle":"2024-06-04T04:08:56.636372Z","shell.execute_reply.started":"2024-06-04T04:08:56.624708Z","shell.execute_reply":"2024-06-04T04:08:56.635182Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_val_vis = np.array(X_val).reshape(X_val.shape[0], 28, 28)\nX_val_vis = X_val_vis\nfig, axis = plt.subplots(int(len(n)/8), 8, figsize=(35, 15))\nfor i, ax in enumerate(axis.flat):\n ax.imshow(X_val_vis[n[i]], cmap='binary')\n digit = np.argmax(y_pred,1)[n[i]]\n digit2 = np.argmax(Y_val,1)[n[i]]\n ax.set(title = f\"Pred Number is {digit}\\n Real Number is {digit2}\");","metadata":{"_uuid":"470a60fd-078e-4a6b-99bf-206ffd71cfb8","_cell_guid":"33ee870d-1eaa-4702-ac0b-7d41a9333040","execution":{"iopub.status.busy":"2024-06-04T04:08:56.638500Z","iopub.execute_input":"2024-06-04T04:08:56.639227Z","iopub.status.idle":"2024-06-04T04:16:00.006782Z","shell.execute_reply.started":"2024-06-04T04:08:56.639193Z","shell.execute_reply":"2024-06-04T04:16:00.004571Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"It looks like even human can't get the right answers.😅","metadata":{"_uuid":"7b0368af-6a8a-44d9-8dee-e994819269d4","_cell_guid":"11579e1a-7889-4904-ab0f-b96007a7069b","trusted":true}},{"cell_type":"markdown","source":"
\n
\n

\nSubmit to Leaderboard\n

\n
\n
","metadata":{"_uuid":"b052fb30-6e45-4c66-87af-ce25efdbe87e","_cell_guid":"739fd7c4-6753-4052-8204-a0038c866f87","trusted":true}},{"cell_type":"code","source":"X_test = np.array(df_test).reshape(-1,28,28,1)","metadata":{"execution":{"iopub.status.busy":"2024-06-04T04:16:00.008574Z","iopub.execute_input":"2024-06-04T04:16:00.008979Z","iopub.status.idle":"2024-06-04T04:16:00.246287Z","shell.execute_reply.started":"2024-06-04T04:16:00.008947Z","shell.execute_reply":"2024-06-04T04:16:00.244441Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = load_model('./best_model.h5')\ny_pred = model.predict(X_test)\ny_pred = np.argmax(y_pred,1)\npd_submit = pd.read_csv('../input/digit-recognizer/sample_submission.csv')\npd_submit['Label'] = y_pred\npd_submit.to_csv('submission.csv', header=True, index=False)","metadata":{"execution":{"iopub.status.busy":"2024-06-04T04:16:00.248697Z","iopub.execute_input":"2024-06-04T04:16:00.249140Z","iopub.status.idle":"2024-06-04T04:16:26.828020Z","shell.execute_reply.started":"2024-06-04T04:16:00.249088Z","shell.execute_reply":"2024-06-04T04:16:26.826733Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# 📃 References 📃","metadata":{"_uuid":"dc3e2507-cb68-47fb-b3e6-3a0c9c7f84f7","_cell_guid":"e301b5a4-ce01-4ba9-ac3b-40983ec40265","trusted":true}},{"cell_type":"markdown","source":"
\n \n
","metadata":{"_uuid":"a9ad3adf-fc21-4ebc-9eb2-d83f80966629","_cell_guid":"f861cd0f-8e0d-48d3-95f9-029e6cf80e84","trusted":true}}]}