{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# **NEURAL NETWORK USING STOCHASTIC GRADIENT DESCENT**\nby **Jean Carlo Codogno** - 03/01/2018\n\n<br/>\n\n+ **INTRODUCTION**\n    + **DATA EXPLORATION**\n        + IMPORT THE DATASET\n        + VISUALIZATION OF THE DATA\n        + COUNT OF EXAMPLES PER DIGIT\n        + NORMALIZATION\n        + ONE HOT ENCODING\n    + **DESCRIBE THE MODEL**\n        + ACTIVATION FUNCTIONS\n        + DESCRIBE THE MODEL\n            + CREATING WEIGHTS\n            + DROPOUT\n            + PREDICT\n            +  METRICS\n            + CROSS - VALIDATION\n\n    + **TRAINING**\n        + SGD (Stochastic Gradient Descent)\n        + MOMENTUM\n        + L2 REGULIZER\n\n\n## INTRODUCTION\n\nThis is a simple Fully Connected  Neural Network using Stochastic Gradient Descent to train it. We use 2 hidden layer with activation function relu and 1 output layer with activation function softmax for classification. \n\n**For computional reason was used a little epochs, in my tests I used 30 epochs and I got 0.98971 in accuracy on public test **\n\n","metadata":{"_uuid":"e857dc986846832b437e58fc5d4d4afe0719a2b2","_execution_state":"busy","_cell_guid":"e2056289-5dea-4153-8f67-539b7fa44b8d"}},{"cell_type":"markdown","source":"## IMPORT LIBRARIES\n\nIt was used the Numpy, Pandas, Sys, Matplotlib and Time.","metadata":{"_uuid":"5db276df1784a107a0e6dad7c1c3d87a73bd6c95","_cell_guid":"4d8eea3f-6478-4815-ad3e-f795fdebb4e8"}},{"cell_type":"code","source":"import matplotlib.pyplot as plt\nimport pandas as pd\nimport numpy as np\nimport time\nimport sys\nimport warnings\nwarnings.filterwarnings(\"ignore\")\n\nseed = 782\nnp.random.seed(seed)","metadata":{"_uuid":"7bcae0a7c40933c487ffa53fd08aabc8071b20ce","_execution_state":"idle","_active":false,"_cell_guid":"86990557-b034-c88c-bc3e-0e0bb666606b","execution":{"iopub.status.busy":"2022-02-08T09:36:55.022509Z","iopub.execute_input":"2022-02-08T09:36:55.022804Z","iopub.status.idle":"2022-02-08T09:36:55.031999Z","shell.execute_reply.started":"2022-02-08T09:36:55.02275Z","shell.execute_reply":"2022-02-08T09:36:55.030811Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## DATA EXPLORATION\n\nThe dataset of Digit Recognizer Competition is the famous **MNIST** (The dataset of handwritten digits), it can be found out at <a style='color:#003d99'>http://yann.lecun.com/exdb/mnist/</a>. This dataset consist of gray-scale images of hand-drawn digits, from zero through nine with shape 28x28 (or 784 pixels) and position in the center. \n\nIn the competition is divided of the dataset in the two parts, train-set (train.csv), test-set (test.csv). In the  train-set contains 42000 examples and in the test-set contain 28000 images to classify.\n\n### IMPORT THE DATASET\n\nIt was used the Pandas to import the data set and transform in numpy array.","metadata":{"_uuid":"d8ef230c63ec1bf19521a234ccd9ef71e449f8a4","_cell_guid":"4d4f088e-6272-4959-a1cf-22d6e7a8d7a1"}},{"cell_type":"code","source":"df = pd.read_csv(\"../input/train.csv\")\ntrain = df.as_matrix()\n\ntrain_y = train[:,0].astype('int8')\ntrain_x = train[:,1:].astype('float64')\n\ntrain = None\n\nprint(\"Shape Train Images: (%d,%d)\" % train_x.shape)\nprint(\"Shape Labels: (%d)\" % train_y.shape)","metadata":{"_uuid":"6c7855986fa68686cc3ef7317881a7d4b023ca0c","_execution_state":"idle","_active":false,"_cell_guid":"48c61204-4638-a851-06da-2e07208e7b35","execution":{"iopub.status.busy":"2022-02-08T09:36:55.038433Z","iopub.execute_input":"2022-02-08T09:36:55.038807Z","iopub.status.idle":"2022-02-08T09:36:59.471242Z","shell.execute_reply.started":"2022-02-08T09:36:55.038736Z","shell.execute_reply":"2022-02-08T09:36:59.470468Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df = pd.read_csv(\"../input/test.csv\")\ntest = df.as_matrix().astype('float64')\nprint(\"Shape Test Images: (%d,%d)\" % train_x.shape)","metadata":{"_uuid":"4cbffbc9c03709c5fe1b229d6a332c9d227caba4","_cell_guid":"2662e464-988d-45b8-8de7-f3e3c7eba585","execution":{"iopub.status.busy":"2022-02-08T09:36:59.472513Z","iopub.execute_input":"2022-02-08T09:36:59.473035Z","iopub.status.idle":"2022-02-08T09:37:01.842568Z","shell.execute_reply.started":"2022-02-08T09:36:59.472982Z","shell.execute_reply":"2022-02-08T09:37:01.84193Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### VISUALIZATION OF THE DATA","metadata":{"_uuid":"0f58798b586ec794853cfe0631c9a77cf461050a","_cell_guid":"79a5a0de-b7ff-41ef-96fb-35bb9f37e757"}},{"cell_type":"code","source":"def show_image(image, shape, label=\"\", cmp=None):\n    img = np.reshape(image,shape)\n    plt.imshow(img,cmap=cmp, interpolation='none')\n    plt.title(label)\n    \n","metadata":{"_uuid":"10cdbbaf8917f96b6bc548c4ec668945ad98aa3e","_cell_guid":"e51e694e-6839-49a3-9aa7-fc1e5fdc7910","execution":{"iopub.status.busy":"2022-02-08T09:37:01.843809Z","iopub.execute_input":"2022-02-08T09:37:01.844095Z","iopub.status.idle":"2022-02-08T09:37:01.850766Z","shell.execute_reply.started":"2022-02-08T09:37:01.844048Z","shell.execute_reply":"2022-02-08T09:37:01.849896Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%matplotlib inline\nplt.figure(figsize=(12,10))\n\ny, x = 5,10\nfor i in range(0,(y*x)):\n    plt.subplot(y, x, i+1)\n    ni = np.random.randint(0,train_x.shape[0],1)[0]\n    show_image(train_x[ni],(28,28), train_y[ni], cmp=\"gray\")\nplt.show()","metadata":{"_uuid":"089462b82caf6ca1719ecc18ead873807db3007a","_cell_guid":"67d33453-f823-40db-8bad-1edb52e4481b","execution":{"iopub.status.busy":"2022-02-08T09:37:01.851999Z","iopub.execute_input":"2022-02-08T09:37:01.852457Z","iopub.status.idle":"2022-02-08T09:37:04.645977Z","shell.execute_reply.started":"2022-02-08T09:37:01.852403Z","shell.execute_reply":"2022-02-08T09:37:04.645197Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### COUNT OF EXAMPLES PER DIGIT\n\nCount of number the examples per digit is important, because if the data isn't uniform we can have a bad accuracy. In this dataset the number of examples per digit is uniform, so we won't worry.","metadata":{"_uuid":"2d8430ebbb5c8e4cb55a3d80232727f0feb1f17f","_cell_guid":"bde5f15c-a4be-457e-8c96-bb10c2f7a6b8"}},{"cell_type":"code","source":"def count_exemple_per_digit(exemples):\n    hist = np.ones(10)\n\n    for y in exemples:\n        hist[y] += 1\n\n    colors = []\n    for i in range(10):\n        colors.append(plt.get_cmap('viridis')(np.random.uniform(0.0,1.0,1)[0]))\n\n    bar = plt.bar(np.arange(10), hist, 0.8, color=colors)\n\n    plt.grid()\n    plt.show()\n\ncount_exemple_per_digit(train_y)","metadata":{"_uuid":"ce0e0758f47f20ab991098fae161bddb121500e4","_cell_guid":"61750351-8e8e-4670-9ad7-c0950224346e","execution":{"iopub.status.busy":"2022-02-08T09:37:04.647296Z","iopub.execute_input":"2022-02-08T09:37:04.647841Z","iopub.status.idle":"2022-02-08T09:37:04.812687Z","shell.execute_reply.started":"2022-02-08T09:37:04.647789Z","shell.execute_reply":"2022-02-08T09:37:04.81203Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### NORMALIZATION\n\nThe precision of the computers is increasing, but isn't perfect. If we set 1e9 (1 billion), add 1e-6 to it million (1e6) times, and after subtract 1e9 (billion) from it, mathematically you should get 1 but we get 0.95... It happens because we add a lot of times a small number in a big number, breaking the precision of the computer.","metadata":{"_uuid":"bd838daa280147077be5c9205607678b1ba7ca9a","_cell_guid":"7de22384-9cf2-49a2-9db0-28f23cc9fd09"}},{"cell_type":"code","source":"num = 1e9\n\nfor i in range(1000000):\n    num += 1e-6\n\nnum -= 1e9\n\nprint(num)\n    ","metadata":{"_uuid":"08fa88cb06777c9da5f3efc3c05d4cc25b6c6e2a","scrolled":true,"_cell_guid":"eeea723f-115e-4094-b83e-4c068261100c","execution":{"iopub.status.busy":"2022-02-08T09:37:04.81391Z","iopub.execute_input":"2022-02-08T09:37:04.814341Z","iopub.status.idle":"2022-02-08T09:37:04.950991Z","shell.execute_reply.started":"2022-02-08T09:37:04.814292Z","shell.execute_reply":"2022-02-08T09:37:04.950326Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"It can be a problem, because Neural Network uses a lot of sum. So, we have used normalization techniques for solving it. In general is a good practice normalize of the data with mean zero and a small variance.\n\nFor image exists a very easy normalization, how the values of an image is between 0 - 255, we just divide by 255. With this normalization the values of the images are between 0 - 1.","metadata":{"_uuid":"da8d9e57743f5283d9369c043a81f1f40c54ee91","_cell_guid":"022db2ab-e598-4375-9793-083770c0258b"}},{"cell_type":"code","source":"def normalization(x, mu, sigma):\n    epsilon = 1e-20\n    X_norm = np.zeros_like(x)\n    \n    for i in range(len(x)):\n        for j in range(len(x[i,:])):\n          X_norm[i,j] = (x[i,j]-mu[j])/(sigma[j]+epsilon)\n\n    return [np.array(X_norm), mu, sigma]","metadata":{"_uuid":"52336acb5703acfa3385e23c8a54e5d21b7fccbd","_cell_guid":"7e62a52f-10f9-4587-9672-8b8708551a0e","execution":{"iopub.status.busy":"2022-02-08T09:37:04.952332Z","iopub.execute_input":"2022-02-08T09:37:04.952589Z","iopub.status.idle":"2022-02-08T09:37:04.963845Z","shell.execute_reply.started":"2022-02-08T09:37:04.952545Z","shell.execute_reply":"2022-02-08T09:37:04.962957Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#train_x /=255\n#test /= 255\n\nmu = np.mean(train_x, axis=0)\nsigma = np.max(train_x, axis=0)-np.min(train_x, axis=0)\n\ntrain_x, mu, sigma = normalization(train_x,mu, sigma)\ntest,_,_ = normalization(test, mu, sigma)\n\nprint(\"Min: %.2f\" % np.min(train_x))\nprint(\"Max: %.2f\" % np.max(train_x))","metadata":{"_uuid":"9d98b46d9d70692be12f54ca56901fda9c2b4d38","_execution_state":"idle","_cell_guid":"0e9d8f99-8de1-49de-8e7f-5d2e3d47aad3","execution":{"iopub.status.busy":"2022-02-08T09:37:04.965209Z","iopub.execute_input":"2022-02-08T09:37:04.965692Z","iopub.status.idle":"2022-02-08T09:38:10.416288Z","shell.execute_reply.started":"2022-02-08T09:37:04.965618Z","shell.execute_reply":"2022-02-08T09:38:10.415516Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### ONE HOT ENCODING\n\nOne Hot Encodding or Dummies is a group of the digits when each digit corresponds the propabilite of the class is True. It's very used for classifying. To do it, we'll use the pandas.","metadata":{"_uuid":"a628a8aaacecfcbdd2fa0fa2d065df6c622045c1","_cell_guid":"4736cb8f-29a2-41f1-85b8-1818eba31abf"}},{"cell_type":"code","source":"train_y = pd.get_dummies(train_y).as_matrix()\nprint(train_y[0])","metadata":{"_uuid":"ba67bbb6e2702b054d475e76fe0ff09e711f460b","_cell_guid":"9beaa97d-9291-4f93-bd9d-8eb5a5e8d4bc","execution":{"iopub.status.busy":"2022-02-08T09:38:10.417614Z","iopub.execute_input":"2022-02-08T09:38:10.418134Z","iopub.status.idle":"2022-02-08T09:38:10.429987Z","shell.execute_reply.started":"2022-02-08T09:38:10.418079Z","shell.execute_reply":"2022-02-08T09:38:10.429172Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## DEFINING THE MODEL\n\nArtificial Neural Network or Neural Network(ANN or NN) is a branch of Machine Learning, that it is inspired by the biological neural networks. The NN consist of the conection of artifical neurons divide in layers. The most famous neuron is the perceptron. \n\nThe artificial neuron consist in simulate the function of the biological neuron. The biological neural learn with its connections, if the connections is enough strong the signal is activate. In artificial neurons this feature is representate with variable Weght (wi), that they are changed in the training. These Weights represente the knowlegde of the Neural Networks.\n\nMathematically a artificial neuron can be represented by the sum of product between Input values and its respective weight, as equation bellow.\n\n$$ Z = \\sum_{i=1}^{n} X_i \\cdot W_i + b_i$$ \n<br/>\nIn the artificial neuron contain the activation function, when the Z is apply. Exist a lot of activation functions, most famous are Sigmoid, Tahn, ReLu, LReLu and PReLu.\n\nThe derivative of the Z will be important for train the model.\n\n$$ \\dfrac{dZ}{dW} = X_i$$ \n\n![image.png](attachment:image.png)\n\nFonte: <a style='color:#003d99'>http://www.theprojectspot.com/tutorial-post/introduction-to-artificial-neural-networks-part-1/7</a>","metadata":{"_uuid":"8fd22c0c702144e9518e9680cf1f6de3bb19c9cb","_cell_guid":"50b0b915-089e-473e-b5d8-f8f51961f851"}},{"cell_type":"markdown","source":"### ACTIVATION FUNCTIONS\n\nExist a lot of types of activations functions, but in this model we'll use the ReLu  (Rectifier Linear Function), this function is similar to the functioning of the biological neuron, and allow us to run the model more fast. ","metadata":{"_uuid":"67fc4eba288d2d078186ee4e9b06433ebdcec8bd","_cell_guid":"d2a9f14e-d3c4-4072-9ad2-fc2be3164461"}},{"cell_type":"code","source":"def ReLu(x, derivative=False):\n    if(derivative==False):\n        return x*(x > 0)\n    else:\n        return 1*(x > 0)\n\nx = np.arange(20)-10\nrelu = ReLu(x)\n\nplt.plot(x, relu)\nplt.show()","metadata":{"_uuid":"f9125249f805c751b61f73cccb5b4c110a53d4f6","_cell_guid":"a69789fd-5692-4fdc-82d4-e9d922f7327f","execution":{"iopub.status.busy":"2022-02-08T09:38:10.431427Z","iopub.execute_input":"2022-02-08T09:38:10.432145Z","iopub.status.idle":"2022-02-08T09:38:10.548477Z","shell.execute_reply.started":"2022-02-08T09:38:10.431782Z","shell.execute_reply":"2022-02-08T09:38:10.547698Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"The Softmax function allow us predict the model, because it normalize the data in one hot encoding.\n\n$$ y = \\sigma(x) = \\dfrac{e^x}{\\sum_{i=1}^{N} e^x}  $$\n\nThe derivative of the σ(x) will be important for train the model.\n\n$$ \\dfrac{dy}{dZ} = \\dfrac{d\\sigma(x)}{dZ} = \\dfrac{e^x.\\sum_{i=1}^{N} e^x - e^x. e^x}{[\\sum_{i=1}^{N} e^x]^2}$$\n$$ \\dfrac{d\\sigma(x)}{dZ} = \\dfrac{e^x.(\\sum_{i=1}^{N}  e^x. e^x)}{\\sum_{i=1}^{N} e^x.\\sum_{i=1}^{N} e^x} $$\n$$ \\dfrac{d\\sigma(x)}{dZ} = \\dfrac{e^x}{\\sum_{i=1}^{N} e^x }. \\dfrac{(\\sum_{i=1}^{N}  e^x - e^x)}{\\sum_{i=1}^{N} e^x} $$\n$$ \\dfrac{d\\sigma(x)}{dZ} = \\dfrac{e^x}{\\sum_{i=1}^{N} e^x }. (\\dfrac{(\\sum_{i=1}^{N}  e^x)}{\\sum_{i=1}^{N} e^x}-\\dfrac{e^x}{\\sum_{i=1}^{N} e^x}) $$\n$$ \\dfrac{d\\sigma(x)}{dZ} = y. (1-y) $$\n","metadata":{"_uuid":"79c47e847a7c0fc8e8a20f889d3c98ce873f0e22","_cell_guid":"1e032e5a-506e-4ddb-bb0f-bf7c7caa07d4"}},{"cell_type":"code","source":"def Softmax(x):\n    x -= np.max(x)\n    sm = (np.exp(x).T / np.sum(np.exp(x),axis=1)).T\n    return sm\n\nx = np.arange(20)-10\nsoftmax = Softmax([x])\n\nplt.plot(x, softmax[0])\nplt.show()","metadata":{"_uuid":"87260255c2ed0afa612d441e9f2912182849a195","_cell_guid":"53a31cab-b171-42ef-9512-0942486c3dbb","execution":{"iopub.status.busy":"2022-02-08T09:38:10.549935Z","iopub.execute_input":"2022-02-08T09:38:10.5505Z","iopub.status.idle":"2022-02-08T09:38:10.769248Z","shell.execute_reply.started":"2022-02-08T09:38:10.550448Z","shell.execute_reply":"2022-02-08T09:38:10.768568Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### DESCRIBE THE MODEL\n\nIn this link <a style='color:#003d99'>http://yann.lecun.com/exdb/mnist/</a> we can see some examples of models and yours respectives error in porcent. In this website the bests models are using the Convolution Neural Network, but it isn't the goal this Notebook. \n\nIn this notebook we will use a Neural Network with 3 layer, 800-300 Hidden Units with Softmax Classification, Cross-Entropy as Loss Function and weight decay, as describe by Hinton, unpublished, 2005. Where he got 1,53% porcent error in this Neural Network.","metadata":{"_uuid":"5ab145330437e6e8a3a1b92b4ea4166f71d9cbd6","_cell_guid":"6b58caeb-8531-4fd3-af76-6afb6f0f4ba8"}},{"cell_type":"markdown","source":"####   CREATING WEIGHTS \n<br/>\nTo simplify and helpin the calculus, we are use the matriz representation for representate the weights, inputs and the outputs.\nIn the method below will create the weights and the biases with mean zero and the variance proportional to root of the numbers of inputs. This method use the Numpy to create the ramdom number with normal function. This method return a list with the weights and biases.","metadata":{"_uuid":"4d4f16b84a9fe59552a977297da2ab090ae2ee6d","_cell_guid":"f6392717-1514-4ce1-853a-8937ff5ef2c4"}},{"cell_type":"code","source":"def CreateWeights():\n    ##Initialization of the Weights and the Biases with the random gaussian function with mean zeron, and variance between 1/sqtr(num_inputs_layer)\n    \n    ninputs = 784\n    wl1 = 500 ##Number of neurons in the first layer\n    wl2 = 300 ##Number of neurons in the second layer\n    nclass = 10 ##Numer of the class, in this case it is the number of the digits.\n    \n    #layer1\n    w1 = np.random.normal(0, ninputs**-0.5, [ninputs,wl1])\n    b1 = np.random.normal(0, ninputs**-0.5, [1,wl1])\n    \n    #Layer2\n    w2 = np.random.normal(0, wl1**-0.5, [wl1,wl2])\n    b2 = np.random.normal(0, wl1**-0.5, [1,wl2])\n\n    #Layer3\n    w3 = np.random.normal(0, wl2**-0.5, [wl2,nclass])\n    b3 = np.random.normal(0, wl2**-0.5, [1,nclass])\n    \n    return [w1,w2,w3,b1,b2,b3]","metadata":{"_uuid":"9fad3b5a8167b90b066156bffae8c6f9ab172b1a","_cell_guid":"83226388-8e11-4ff0-ae6b-bcf416a3d759","execution":{"iopub.status.busy":"2022-02-08T09:38:10.770409Z","iopub.execute_input":"2022-02-08T09:38:10.770829Z","iopub.status.idle":"2022-02-08T09:38:10.798476Z","shell.execute_reply.started":"2022-02-08T09:38:10.77077Z","shell.execute_reply":"2022-02-08T09:38:10.797632Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### DROPOUT\n<br/>\nDropout is a simple technique for prevent overfitting, it was approached in the paper \"Dropout: A Simple Way to Prevent Neural Networks from Overfitting\"<a style='color:#003d99'>(https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf)</a>.\nThis technique consist in randomly dropping neurons, for that layer not to depend of these neurons, to predict right.\n![image.png](attachment:image.png)","metadata":{"_uuid":"344866f2d55348d97261dc9ad10d05fb62db8f92","_cell_guid":"4d82c0f5-41c0-4f55-b4f4-638adf51aa51"},"attachments":{"image.png":{"image/png":"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"}}},{"cell_type":"code","source":"def Dropout(x, keep_prop):\n    mask = np.random.binomial([np.ones_like(x)],(1-keep_prop))[0]  / (1-keep_prop)\n    return x*mask","metadata":{"_uuid":"a5382fc411ef29368fadc663e8585960ce81e6f6","_cell_guid":"58f69a54-f2d4-4e28-8634-a857b0fa807b","execution":{"iopub.status.busy":"2022-02-08T09:38:10.802303Z","iopub.execute_input":"2022-02-08T09:38:10.802582Z","iopub.status.idle":"2022-02-08T09:38:10.81054Z","shell.execute_reply.started":"2022-02-08T09:38:10.802528Z","shell.execute_reply":"2022-02-08T09:38:10.809575Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### PREDICT\n<br/>\nThe model will be a Neural Network with 2 layers Fully Connected with  respectively, 500 and 300 neurons, the boths layers with Activation Function  ReLu. The last layer contain 10 neurons and used the Activation Function Softmax. Between the layer we will use the Dropout for prevent overfitting.\n\n> 500 NN - DROPOUT- 300 NN - DROPOUT - 10 NN\n\nHow we are using the matrix representation, we have to use the matrix multiplication. The matrix ","metadata":{"_uuid":"648773855dcd9fbfd1eaabeaef5b5ba8b4728f8a","_cell_guid":"94d1b031-99d6-44da-b6f8-4c32743735a2"}},{"cell_type":"code","source":"def predict(weights, x, keep_prop=0):\n    \n    w1,w2,w3,b1,b2,b3  = weights \n    \n    #1-Hidden Layer\n    first = ReLu(x@w1+b1)\n    first = Dropout(first, keep_prop)\n\n    #2-Hidden Layer\n    second = ReLu(first@w2+b2)\n    second = Dropout(second, keep_prop)\n    \n    #Output Layer\n    return [first, second, Softmax(second@w3+b3)]","metadata":{"_uuid":"796dbeed2aa96c51d93c1776f6c2303f32c7d387","_cell_guid":"8321e098-6d5d-4388-8a49-587829c0799b","execution":{"iopub.status.busy":"2022-02-08T09:38:10.812052Z","iopub.execute_input":"2022-02-08T09:38:10.812693Z","iopub.status.idle":"2022-02-08T09:38:10.829206Z","shell.execute_reply.started":"2022-02-08T09:38:10.812462Z","shell.execute_reply":"2022-02-08T09:38:10.82817Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### METRICS\n<br/>\nAfter to run the model we need to evaluate the model for measure how well the it is going, we can use the accuracy metric and / or loss metric.\nThe accuracy is measure by the percent of the hit of the model did, and we can measure with the function bellow.","metadata":{"_uuid":"3a26e7882025a0ae175173e664918f13e195cf57","_cell_guid":"3289f1ef-070f-492b-9fcd-493511e8f920"}},{"cell_type":"code","source":"def accuracy(output, y):\n    hit = 0\n    output = np.argmax(output, axis=1)\n    y = np.argmax(y, axis=1)\n    for y in zip(output, y):\n        if(y[0]==y[1]):\n            hit += 1\n\n    p = (hit*100)/output.shape[0]\n    return p","metadata":{"_uuid":"23dd420a9a47d17072c6724b869dbf6b05527c26","_execution_state":"idle","_active":false,"_cell_guid":"7f2c05ac-da5c-2513-fde6-27d85ffc3956","execution":{"iopub.status.busy":"2022-02-08T09:38:10.830966Z","iopub.execute_input":"2022-02-08T09:38:10.831686Z","iopub.status.idle":"2022-02-08T09:38:10.846266Z","shell.execute_reply.started":"2022-02-08T09:38:10.831549Z","shell.execute_reply":"2022-02-08T09:38:10.845509Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Exist a lot of loss functions, but as the variance of the numbers of inputs are small the changes will be small also, so we chose the cross-entropy as loss function, because it have high sensibility by small changes. The cross-entropy is used with probability, and compare the outputs of the model with the true results and return the distorcing of the diference.\n\n$$ L(t|y) = - \\dfrac{1}{N}. \\sum_{i=1}^{N} [t \\cdot log(y) + (1-t) \\cdot log(1-y)] $$\n\nWhere, **N** is the number of the exemples, **t** is the right labels and **y** is the predict labels.","metadata":{"_uuid":"00ee870e37c4c2c4b8ed351c3cd3be09f7d54c3c","_cell_guid":"425f1e1b-2586-4525-b742-1f06b1669d28"}},{"cell_type":"code","source":"fig, (ax0, ax1) = plt.subplots(ncols=2, figsize=(14, 4))\n\ny = np.arange(100)/100\n\ne = -np.log(y)\ne2 = -np.log(1-y)\n\nax0.plot(y, e)\nax0.set_title(\"For t (right label) = 1\")\nax0.set_xlabel(\"Predict\")\nax0.set_ylabel(\"Error\")\nax1.plot(y, e2)\nax1.set_title(\"For t (right label) = 0\")\nax1.set_xlabel(\"Predict\")\nax1.set_ylabel(\"Error\")\nplt.show()","metadata":{"_uuid":"8a5b39e242d490e7c4ed88881707d1f3200d207a","_cell_guid":"2cdb61ee-fd90-4943-92dc-bdd5467d9af3","execution":{"iopub.status.busy":"2022-02-08T09:38:10.849594Z","iopub.execute_input":"2022-02-08T09:38:10.85037Z","iopub.status.idle":"2022-02-08T09:38:11.065652Z","shell.execute_reply.started":"2022-02-08T09:38:10.850163Z","shell.execute_reply":"2022-02-08T09:38:11.064868Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"This Loss function works well, but we need to be careful with the **log()**, because the log of the number zeros isn't exist so we consider **log(0)** as the lowest value -inf after use the np.nan_to_num().\n\nThe derivative of the L(t|y) will be important for train the model.\n\n$$ \\dfrac{dL(t|y)}{dy}  = \\dfrac{1}{N}. \\dfrac{t}{y}-\\dfrac{(1-t)}{(1-y)} $$\n$$ = \\dfrac{1}{N}. \\dfrac{t.(1-y) - (1-t).y}{y.(1-y)} $$\n$$ = \\dfrac{1}{N}. \\dfrac{t-ty - y+ty}{y.(1-y)} $$\n$$ = \\dfrac{1}{N}. \\dfrac{t - y}{y.(1-y)} $$\n\nhttp://users.isr.ist.utl.pt/~wurmd/Livros/school/Bishop%20-%20Pattern%20Recognition%20And%20Machine%20Learning%20-%20Springer%20%202006.pdf pg. 206","metadata":{"_uuid":"183b64d0a927e913069b66ace2ed21df33f35014","_cell_guid":"77de75cf-6a35-4f4b-bc5e-69ef4889a241"}},{"cell_type":"code","source":"def log2(x):\n    if(x!=0):\n        return np.log(x)\n    else:\n        return -np.inf\n    \ndef log(y):\n    return [[log2(nx) for nx in x]for x in y]\n\ndef cost(Y_predict, Y_right):\n    \n    Loss = -np.mean(Y_right*log(Y_predict) + (1-Y_right)*log(1-Y_predict))\n    return Loss","metadata":{"_uuid":"d122bb7bb80270688e4ec0b414402d594080b55a","_cell_guid":"1f23f013-60e8-4d71-8110-e0c54381d48d","execution":{"iopub.status.busy":"2022-02-08T09:38:11.066795Z","iopub.execute_input":"2022-02-08T09:38:11.067074Z","iopub.status.idle":"2022-02-08T09:38:11.082445Z","shell.execute_reply.started":"2022-02-08T09:38:11.067012Z","shell.execute_reply":"2022-02-08T09:38:11.081704Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### CROSS - VALIDATION\n<br/>\nThe cross-validation is a techinique used for measure the accuracy and visualization overfitting. This technique consist in split the data in train_data anda test_data, where the test_data consist of 10% - 30% of the all data.","metadata":{"_uuid":"0f5e6ac52d8f135f455a9b2fa9cba91747fe988d","_cell_guid":"579545a6-2bd4-490d-94af-bfe453543eae"}},{"cell_type":"code","source":"porcent_valid = 0.1\nVALID_SIZE = round(train_x.shape[0]*porcent_valid)\n\nindex_data = np.arange(train_x.shape[0])\nnp.random.shuffle(index_data)\n\nx_train = train_x[index_data[VALID_SIZE:]]\nx_valid = train_x[index_data[:VALID_SIZE]]\n\n\nd_train = train_y[index_data[VALID_SIZE:]]\nd_valid = train_y[index_data[:VALID_SIZE]]\n\ntrain_x = None\ntrain_y = None\n\nx_train.shape","metadata":{"_uuid":"6392f94a1182f9e28e6a223f4732bfe98a888d3b","_cell_guid":"f86330c7-79a1-4c65-8a50-9bf74fc3390f","execution":{"iopub.status.busy":"2022-02-08T09:38:11.083738Z","iopub.execute_input":"2022-02-08T09:38:11.084401Z","iopub.status.idle":"2022-02-08T09:38:12.019547Z","shell.execute_reply.started":"2022-02-08T09:38:11.084195Z","shell.execute_reply":"2022-02-08T09:38:12.018921Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## TRAINING\n### SGD (Stochastic Gradient Descent)\n\nSGD is a optimizer used for fit the neural network, this technique is based by Gradient Descent.\nIn SGD is used the matriz representation, the equation for represent the update the weights is bellow.\n\n$$ V_{k+1} = V_k - \\eta . \\nabla L(W_{ij}) $$\n<br/>\n$$  W = V_{k+1}  $$\n\nWhere the **η** is the step size, and the **∇L(W)** is the gradient of the Loss.\n\nThe gradient can be solved using the chain rule of the derivate the Loss funtion by Weights. The equation for output layer is bellow.\n\n$$ \\nabla L(W_{ij}) = \\dfrac{dL(W_{ij})}{dW_{ij}} $$\n<br/>\n$$ =\\dfrac{dL(t|y)}{dy}.\\dfrac{dy}{dZ}.\\dfrac{dZ}{dW_{ij}} $$\n<br/>\n$$ =\\dfrac{1}{N} . \\dfrac{(t-y)}{y.(1-y)}.y.(1-y). X_{ij} $$\n<br/>\n$$ = \\dfrac{1}{N} . (t-y)X_{ij} $$\n<br/>\n<center>For the hidden layer:</center>\n<br/>\n$$ \\nabla L(W_{ij}) = \\dfrac{dL(W_{ij})}{dW_{ij}} $$\n<br/>\n$$ =\\dfrac{dL(t|y_{ij})}{dy_{ij}}.\\dfrac{dy_{ij}}{dZ_{ij}}.\\dfrac{dZ_{ij}}{dy_{ij}}.\\dfrac{dy_{ij}}{dZ_{ij}}.\\dfrac{dZ_{ij}}{dW_{ij}} $$\n<br/>\n$$ = \\dfrac{1}{N} . \\dfrac{(t-y)}{y.(1-y)}.y.(1-y). W_{ij}.\\sigma(x)'.X_{ij} $$\n<br/>\n$$ = \\dfrac{1}{N} . (t-y).W_{ij}.\\sigma(x)'.X_{ij} $$\n<br/>\n### MOMENTUM\n\nThe SGD in training have a lot of oscillations, so the momentum term was invented. The momentum term is used for soft the oscillations and accelerates of the convergence.\n\n$$  V_{i+1} = \\gamma V_i + \\eta . \\nabla L(W) $$ \n<br/>\n$$  W = W - V_{i+1} $$\n<br/>\n\n### L2 REGULIZER \n\nThe L2 regulizer is the technique which is used for penalize the higher weights, \n\n$$ L(t|y|w) = -\\dfrac{1}{N} . \\sum_{i=1}^{N} [(t \\cdot log(y) + (1-t) \\cdot log(1-y)) +\\dfrac{\\lambda}{2}. \\sum_{j=1}^{nj}\\sum_{i=1}^{ni} W_{ij}^2] $$\n\nWith this modification in the Loss function the equation of the update the weights is changed. \nFor the output layer.\n \n$$ \\nabla L(W_{ij}) = \\dfrac{1}{N} . [(t-y).X_{ij} + \\lambda W_{ij}] $$\n\nFor the hidden layer.\n\n$$ \\nabla L(W_{ij}) = \\dfrac{1}{N} . [(t-y).W_{ij}.\\sigma(x)'.X_{ij} +  \\lambda W_{ij} ] $$","metadata":{"_uuid":"673366714707df9b34eff7ef62cab729512c136d","_cell_guid":"58719bf3-d2cf-4fc2-85f0-eccf813c2a7f"}},{"cell_type":"code","source":"def SGD(weights, x, t, outputs, eta, gamma, nabla, cache=None):\n    \n    w1,w2,w3,b1,b2,b3  = weights\n    \n    \n    if(cache==None):\n            vw1 = np.zeros_like(w1)\n            vw2 = np.zeros_like(w2)\n            vw3 = np.zeros_like(w3)\n            vb1 = np.zeros_like(b1)\n            vb2 = np.zeros_like(b2)\n            vb3 = np.zeros_like(b3)\n    else:\n        vw1,vw2,vw3,vb1,vb2,vb3 = cache\n    \n    first, second, y = outputs\n   \n    w3_delta = (t-y)\n    \n    w2_error = w3_delta@w3.T\n\n    w2_delta = w2_error * ReLu(second,derivative=True)\n\n    w1_error = w2_delta@w2.T\n    w1_delta = w1_error * ReLu(first,derivative=True)\n    \n    eta = -eta/x.shape[0]\n \n    vw3 = gamma*vw3 + eta * (second.T@w3_delta + nabla*w3)\n    vb3 = gamma*vb3 + eta * w3_delta.sum(axis=0)\n\n    vw2 = gamma*vw2 + eta * (first.T@w2_delta + nabla*w2)\n    vb2 = gamma*vb2 + eta * w2_delta.sum(axis=0)\n\n    vw1 = gamma*vw1 + eta * (x.T@w1_delta + nabla*w1)\n    vb1 = gamma*vb1 + eta * w1_delta.sum(axis=0)\n    \n    \n    w3 -= vw3\n    b3 -= vb3\n\n    w2 -= vw2\n    b2 -= vb2\n\n    w1 -= vw1\n    b1 -= vb1\n    \n    weights = [w1,w2,w3,b1,b2,b3]\n    cache = [vw1,vw2,vw3,vb1,vb2,vb3]\n    \n    return weights, cache","metadata":{"_uuid":"25d47c0dad5d6a9bf2988963bcab72e9ed388101","_execution_state":"idle","_cell_guid":"239358a6-b413-41db-b49d-34098edd58d9","execution":{"iopub.status.busy":"2022-02-08T09:38:12.020813Z","iopub.execute_input":"2022-02-08T09:38:12.021081Z","iopub.status.idle":"2022-02-08T09:38:12.12844Z","shell.execute_reply.started":"2022-02-08T09:38:12.021038Z","shell.execute_reply":"2022-02-08T09:38:12.127701Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Training process\n\nIn the training process is divided all the data in batchs, it's a process part because it's principle fundament of the stochastic process. These batchs need to be divide with randomlly for improved the accuracy of the model.\n\n### Image Transformation\n\nSimard, Steinkraus and Platt in yours paper \"Best Practices for Convolutional Neural Networks applied to Visual Document Analysis\"(http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.10.5032&rep=rep1&type=pdf) show that the transformation images can improved the Neural Network. In yours tests the Elastic Transformation showed the best results.\n\nI found out in the github a method that implement the elastic transformation. (https://gist.github.com/fmder/e28813c1e8721830ff9c)","metadata":{"_uuid":"6d7bdd4407ded45dd9697233eb438eae7a8b84c8","_cell_guid":"e89e4b01-0df5-4a78-9fa3-9363168ee76e"}},{"cell_type":"code","source":"from scipy.ndimage.interpolation import map_coordinates\nfrom scipy.ndimage.filters import gaussian_filter\n\ndef elastic_transform(image, alpha, sigma, random_state=None):\n    \"\"\"Elastic deformation of images as described in [Simard2003]_.\n    .. [Simard2003] Simard, Steinkraus and Platt, \"Best Practices for\n       Convolutional Neural Networks applied to Visual Document Analysis\", in\n       Proc. of the International Conference on Document Analysis and\n       Recognition, 2003.\n    \"\"\"\n    if random_state is None:\n        random_state = np.random.RandomState(None)\n\n    shape = image.shape\n    dx = gaussian_filter((random_state.rand(*shape) * 2 - 1), sigma, mode=\"constant\", cval=0) * alpha\n    dy = gaussian_filter((random_state.rand(*shape) * 2 - 1), sigma, mode=\"constant\", cval=0) * alpha\n\n    x, y = np.meshgrid(np.arange(shape[0]), np.arange(shape[1]))\n    indices = np.reshape(y+dy, (-1, 1)), np.reshape(x+dx, (-1, 1))\n\n    return map_coordinates(image, indices, order=1).reshape(shape)","metadata":{"_uuid":"5747a19626f4afe6f4f5af11e57467679b6ed544","_cell_guid":"016426fd-ab73-4e4c-ad2a-8703c3c91bb3","execution":{"iopub.status.busy":"2022-02-08T09:38:12.131183Z","iopub.execute_input":"2022-02-08T09:38:12.131872Z","iopub.status.idle":"2022-02-08T09:38:12.55813Z","shell.execute_reply.started":"2022-02-08T09:38:12.131815Z","shell.execute_reply":"2022-02-08T09:38:12.556978Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"x_t = np.array([elastic_transform(xx.reshape(28,28),15,3).reshape(784) for xx in x_train[0:10]])\nplt.subplot(1,2,1)\nplt.imshow(x_t[0].reshape(28,28))\nplt.subplot(1,2,2)\nplt.imshow(x_train[0].reshape(28,28))","metadata":{"_uuid":"7bbc3ddcf83eb9003185d9fd8df72ad77ef283f1","_cell_guid":"f78f9879-9e87-4476-ab42-6420300eacd2","execution":{"iopub.status.busy":"2022-02-08T09:38:12.564581Z","iopub.execute_input":"2022-02-08T09:38:12.564971Z","iopub.status.idle":"2022-02-08T09:38:12.871809Z","shell.execute_reply.started":"2022-02-08T09:38:12.564877Z","shell.execute_reply":"2022-02-08T09:38:12.870997Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\ndef run(weights, x_train, y_train, x_valid, y_valid, epochs = 10, nbatchs=25, alpha = 1e-3, decay = 0, momentum = 0, l2 = 0.001, keep_prop = 0):\n    \n    pross = x_train.shape[0]*0.05\n    \n    history = [[],[]]\n    \n    index = np.arange(x_train.shape[0])\n    cache = None\n    print(\"Train data: %d\" % (x_train.shape[0]))\n    print(\"Validation data: %d \\n\" % (x_valid.shape[0]))\n    mtime = 0\n    \n    for j in range(epochs):\n        np.random.shuffle(index)\n        t = 0\n        iterations = round(x_train.shape[0]/nbatchs)\n        prog = \"\"\n        sacurr = 0\n        sloss = 0\n        sys.stdout.write(\"\\nEpochs: %2d \\ %2d \\n\"% (j+1,epochs))\n        stime = 0\n        timeIT = time.time()\n        for i in range(iterations):\n            timeI = time.time()\n            f = i*nbatchs\n            l = f+nbatchs\n            \n            if(l>(x_train.shape[0]-1)):\n                l = x_train.shape[0]\n                \n            x = np.array([elastic_transform(xx.reshape(28,28),15,3).reshape(784) for xx in x_train[index[f:l]]])\n            y = y_train[index[f:l]]\n\n            outputs = predict(weights, x, keep_prop)\n            \n            loss = cost(outputs[-1], y)\n            \n            \n            accuracy_t = accuracy(outputs[-1], y)\n            \n            sacurr += accuracy_t\n            sloss += loss\n            \n            accuracy_train = sacurr/(i+1)\n            loss_train = sloss/(i+1)\n            \n            weights, cache = SGD(weights, x, y, outputs, alpha, momentum, l2, cache)\n            \n            t+= x.shape[0]\n            \n            qtd = round(t/pross)\n            prog = \"[\"\n            for p in range(20):\n                if(p<qtd-1):\n                    prog += \"=\"\n                elif(p==qtd-1):\n                    prog += \">\"\n                else:\n                    prog += \".\"\n            prog += \"]\"\n\n            \n            stime += time.time()-timeI\n            mtime = stime/(i+1)\n            mTimeT = mtime * (iterations-i-1)\n            \n            sys.stdout.write(\"\\r%5d/%5d %s ETA: %3d s - loss: %.4f  acc: %.4f\" % (t, x_train.shape[0], prog, mTimeT, loss_train, accuracy_train))\n            \n            history[0].append([loss_train, accuracy_train])\n        mtime = time.time()-timeIT\n        alpha = alpha - (alpha*decay)\n        \n        outputs = predict(weights, x_valid)\n        \n        loss_valid = cost(outputs[-1], y_valid)\n        accuracy_valid = accuracy(outputs[-1], y_valid)\n        \n        sys.stdout.write(\"\\r%5d/%5d %s ETA: %3d s loss: %.4f  acc: %.4f - lossValid: %.4f  accValid: %.4f \" % ( t, x_train.shape[0], prog, mtime, loss_train, accuracy_train, loss_valid, accuracy_valid))\n        history[1].append([loss_valid, accuracy_valid])\n        \n    return weights, history","metadata":{"_uuid":"be8bea97cddcadbc2ee018101373ff2937943a88","_execution_state":"idle","_active":false,"_cell_guid":"c421dda0-0e37-9d62-8a8d-7660befa4bdd","execution":{"iopub.status.busy":"2022-02-08T09:38:12.876419Z","iopub.execute_input":"2022-02-08T09:38:12.876804Z","iopub.status.idle":"2022-02-08T09:38:13.2552Z","shell.execute_reply.started":"2022-02-08T09:38:12.876677Z","shell.execute_reply":"2022-02-08T09:38:13.254392Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"weights = CreateWeights()\n\nalpha = 5e-2\nepochs = 40\nnbatchs = 100\nweights, history = run(weights, \n              x_train, d_train, \n              x_valid, d_valid, \n              epochs = epochs,\n              nbatchs=nbatchs, \n              alpha = alpha, \n              decay = 0.05, \n              momentum = 0.9, \n              l2 = 1e-7, \n              keep_prop = 0.25)","metadata":{"_uuid":"5cb3502552ff369cb0870625d4c11cd5282bfc7a","_execution_state":"idle","_cell_guid":"8935ff94-fd2a-4e19-9d4e-c31a47fda938","scrolled":true,"execution":{"iopub.status.busy":"2022-02-08T09:39:10.107579Z","iopub.execute_input":"2022-02-08T09:39:10.107854Z","iopub.status.idle":"2022-02-08T09:40:15.8457Z","shell.execute_reply.started":"2022-02-08T09:39:10.107799Z","shell.execute_reply":"2022-02-08T09:40:15.844102Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_history = np.array(history[0])\nt_loss = train_history[:,:1]\nt_acc = train_history[:,1:2]","metadata":{"_uuid":"71791cb7135886bb2dacb34c78925efb278e4fcb","_cell_guid":"ad2396a3-ebe5-46f9-9d26-3e097e4a449d","execution":{"iopub.status.busy":"2022-02-08T09:40:21.217001Z","iopub.execute_input":"2022-02-08T09:40:21.217366Z","iopub.status.idle":"2022-02-08T09:40:21.224406Z","shell.execute_reply.started":"2022-02-08T09:40:21.217298Z","shell.execute_reply":"2022-02-08T09:40:21.223637Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"valid_history = np.array(history[1])\ntrain_history.shape\nv_loss = valid_history[:,:1]\nv_acc = valid_history[:,1:2]","metadata":{"_uuid":"b80b3eae0e0dc66a73547c6e5bbba30203730e42","_cell_guid":"614bc2a5-a316-47b5-a468-8f5524a354df","execution":{"iopub.status.busy":"2022-02-08T09:40:21.225742Z","iopub.execute_input":"2022-02-08T09:40:21.22636Z","iopub.status.idle":"2022-02-08T09:40:21.235565Z","shell.execute_reply.started":"2022-02-08T09:40:21.226281Z","shell.execute_reply":"2022-02-08T09:40:21.234704Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(12,10))\n\nplt.subplot(2, 1, 1)\nx = np.arange(epochs)*int(x_train.shape[0]/nbatchs)\nplt.plot(x,v_acc)\nplt.plot(t_acc)\n\nplt.subplot(2, 1, 2)\nplt.plot(x,v_loss)\nplt.plot(t_loss)\n\nplt.show()","metadata":{"_uuid":"7b391d16c2f9d089188d95f0b55591fa454cf403","_cell_guid":"ef9209f3-797e-42d6-bea2-893d6ef65f7a","execution":{"iopub.status.busy":"2022-02-08T09:40:21.237027Z","iopub.execute_input":"2022-02-08T09:40:21.237539Z","iopub.status.idle":"2022-02-08T09:40:21.477861Z","shell.execute_reply.started":"2022-02-08T09:40:21.237483Z","shell.execute_reply":"2022-02-08T09:40:21.477093Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def plot_confusion_matrix(cm, classes,\n                          normalize=False,\n                          title='Confusion matrix',\n                          cmap=plt.cm.Blues):\n    \"\"\"\n    This function prints and plots the confusion matrix.\n    Normalization can be applied by setting `normalize=True`.\n    \"\"\"\n    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n    plt.title(title)\n    plt.colorbar()\n    tick_marks = np.arange(len(classes))\n    plt.xticks(tick_marks, classes, rotation=45)\n    plt.yticks(tick_marks, classes)\n\n    if normalize:\n        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n\n    thresh = cm.max() / 2.\n    for i in range(cm.shape[0]):\n        for j in range(cm.shape[1]):\n            plt.text(j, i, cm[i, j],\n                     horizontalalignment=\"center\",\n                     color=\"white\" if cm[i, j] > thresh else \"black\")\n\n    plt.tight_layout()\n    plt.ylabel('True label')\n    plt.xlabel('Predicted label')","metadata":{"_uuid":"211321cf0c18e8d90f2e4f3f950b46f9054e0cf4","_cell_guid":"2536ea72-bc6d-4ecb-9129-90ea5187605d","execution":{"iopub.status.busy":"2022-02-08T09:40:21.479205Z","iopub.execute_input":"2022-02-08T09:40:21.479634Z","iopub.status.idle":"2022-02-08T09:40:21.514784Z","shell.execute_reply.started":"2022-02-08T09:40:21.479584Z","shell.execute_reply":"2022-02-08T09:40:21.514174Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"outputs = predict(weights, x_valid)","metadata":{"_uuid":"399eb8da2804358c64f6196f8ca4fe8c7e1f5667","_cell_guid":"082ab351-e29a-4af8-bf81-04ea38378885","execution":{"iopub.status.busy":"2022-02-08T09:40:21.516315Z","iopub.execute_input":"2022-02-08T09:40:21.516873Z","iopub.status.idle":"2022-02-08T09:40:21.712284Z","shell.execute_reply.started":"2022-02-08T09:40:21.516803Z","shell.execute_reply":"2022-02-08T09:40:21.7115Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"p = np.argmax(outputs[-1],axis=1)\npredict_dummies = pd.get_dummies(p).as_matrix().astype('int8')\n\ncm = np.zeros((10,10)).astype(np.int64)\n\nd_valid_int = np.argmax(d_valid, axis=1)\n\nfor i in range(predict_dummies.shape[0]):\n    cm[d_valid_int[i]] += predict_dummies[i]\n\nprint(cm)","metadata":{"_uuid":"ccb608f239c49f4387c0e27470d23af2b302006c","_cell_guid":"cd8797da-5a85-4220-9a21-f75f7c83ddad","execution":{"iopub.status.busy":"2022-02-08T09:40:21.715913Z","iopub.execute_input":"2022-02-08T09:40:21.716198Z","iopub.status.idle":"2022-02-08T09:40:21.754729Z","shell.execute_reply.started":"2022-02-08T09:40:21.716151Z","shell.execute_reply":"2022-02-08T09:40:21.753273Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_confusion_matrix(cm, range(10))","metadata":{"_uuid":"686c0ef9f9fc15fdb3da3ce5ab172e21b038ee17","_cell_guid":"ad306933-a0e8-4b1c-89c2-fefa324b439f","execution":{"iopub.status.busy":"2022-02-08T09:40:21.756801Z","iopub.execute_input":"2022-02-08T09:40:21.757285Z","iopub.status.idle":"2022-02-08T09:40:22.149627Z","shell.execute_reply.started":"2022-02-08T09:40:21.757236Z","shell.execute_reply":"2022-02-08T09:40:22.148861Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"s_cm = np.sum(cm,axis=0)\nfor i in range(10):\n    p = cm[i][i]/s_cm[i]\n    print(\"%d - %.3f %%\" % (i,p))\n    ","metadata":{"_uuid":"ef26d0b4908ff4d6ab8cfeaa6b03bdf67ac958af","_cell_guid":"ccb5eae4-550d-4948-afbc-c19a2ebc9243","execution":{"iopub.status.busy":"2022-02-08T09:40:22.15087Z","iopub.execute_input":"2022-02-08T09:40:22.151374Z","iopub.status.idle":"2022-02-08T09:40:22.162766Z","shell.execute_reply.started":"2022-02-08T09:40:22.151308Z","shell.execute_reply":"2022-02-08T09:40:22.161607Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"w1,w2,w3,b1,b2,b3  = weights","metadata":{"_uuid":"ae65308113e168f8c59963271dec4720a8b07ef7","_execution_state":"idle","_cell_guid":"fff7a62f-1b77-4f1a-8579-4f44fb5c469c","execution":{"iopub.status.busy":"2022-02-08T09:40:22.164579Z","iopub.execute_input":"2022-02-08T09:40:22.165009Z","iopub.status.idle":"2022-02-08T09:40:22.170387Z","shell.execute_reply.started":"2022-02-08T09:40:22.164921Z","shell.execute_reply":"2022-02-08T09:40:22.169443Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%matplotlib inline\nplt.figure(figsize=(12,10))\n\ny, x = 4,10\nfor i in range(0,(y*x)):\n    ni = np.random.randint(0,w1.shape[1],1)[0]\n    plt.subplot(y, x, i+1)\n    plt.imshow(w1[:,ni].reshape((28,28)))\n\nplt.show()","metadata":{"_uuid":"90f3a35e646c16048876584a7ba1e99213393c87","_execution_state":"idle","_cell_guid":"eb64115d-6e72-44e2-ad9b-ddc5495af573","scrolled":true,"execution":{"iopub.status.busy":"2022-02-08T09:40:22.171854Z","iopub.execute_input":"2022-02-08T09:40:22.172554Z","iopub.status.idle":"2022-02-08T09:40:24.297493Z","shell.execute_reply.started":"2022-02-08T09:40:22.172402Z","shell.execute_reply":"2022-02-08T09:40:24.296839Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%matplotlib inline\nplt.figure(figsize=(12,10))\n\ny, x = 4,10\nfor i in range(0,(y*x)):\n    ni = np.random.randint(0,w2.shape[1],1)[0]\n    plt.subplot(y, x, i+1)\n    plt.imshow(w2[:,ni].reshape((20,25)))\n\nplt.show()","metadata":{"_uuid":"f9d1e571519c9e144dea458d8c7766f4c4630305","_execution_state":"idle","_cell_guid":"156fa5f4-99fb-41f2-8b5f-e7ea1f7554da","execution":{"iopub.status.busy":"2022-02-08T09:40:24.298518Z","iopub.execute_input":"2022-02-08T09:40:24.298759Z","iopub.status.idle":"2022-02-08T09:40:26.763981Z","shell.execute_reply.started":"2022-02-08T09:40:24.298721Z","shell.execute_reply":"2022-02-08T09:40:26.763233Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%matplotlib inline\nplt.figure(figsize=(12,10))\n\ny, x = 1,10\nfor i in range(0,(y*x)):\n    plt.subplot(y, x, i+1)\n    plt.imshow(w3[:,i].reshape((20,15)))\n\nplt.show()","metadata":{"_uuid":"6cfb3e3a8869a9e2f74ca63a3ac9c62ebdde2ad0","_execution_state":"idle","_cell_guid":"5e1b7099-c480-4ce6-a83d-e21b8932c437","execution":{"iopub.status.busy":"2022-02-08T09:40:26.765357Z","iopub.execute_input":"2022-02-08T09:40:26.765645Z","iopub.status.idle":"2022-02-08T09:40:27.307054Z","shell.execute_reply.started":"2022-02-08T09:40:26.765582Z","shell.execute_reply":"2022-02-08T09:40:27.306309Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"outputs = predict(weights, test)\n","metadata":{"_uuid":"3e9bc190fb7f15b43a449600ee77f9be7c4deaba","_execution_state":"idle","_cell_guid":"751a050e-2838-4b8f-a528-c49d50e585ff","execution":{"iopub.status.busy":"2022-02-08T09:40:27.30819Z","iopub.execute_input":"2022-02-08T09:40:27.308599Z","iopub.status.idle":"2022-02-08T09:40:28.717838Z","shell.execute_reply.started":"2022-02-08T09:40:27.308546Z","shell.execute_reply":"2022-02-08T09:40:28.717171Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"d = np.argmax(outputs[-1],axis=1)","metadata":{"_uuid":"d804ca60bee49063a1875eb57e4d966f80e23b68","_execution_state":"idle","_cell_guid":"1bd72876-518a-44ff-a4d8-b6655bb043f0","execution":{"iopub.status.busy":"2022-02-08T09:40:28.719153Z","iopub.execute_input":"2022-02-08T09:40:28.71943Z","iopub.status.idle":"2022-02-08T09:40:28.724003Z","shell.execute_reply.started":"2022-02-08T09:40:28.719386Z","shell.execute_reply":"2022-02-08T09:40:28.723258Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"%matplotlib inline\nplt.figure(figsize=(20,12))\n\ny, x = 5,11\nfor i in range(0,(y*x)):\n    plt.subplot(y, x, i+1)\n    ni = np.random.randint(0,test.shape[0],1)[0]\n    v = str(d[ni])\n    show_image(test[ni],(28,28), v, cmp=\"gray\")\nplt.show()","metadata":{"_uuid":"e2a3dbcaaadd8a3c85a0a51d8f75da8b721507cf","_execution_state":"idle","_cell_guid":"d5131c97-26b6-41bc-b8bb-440c4353df67","execution":{"iopub.status.busy":"2022-02-08T09:40:28.725773Z","iopub.execute_input":"2022-02-08T09:40:28.726278Z","iopub.status.idle":"2022-02-08T09:40:31.79945Z","shell.execute_reply.started":"2022-02-08T09:40:28.726139Z","shell.execute_reply":"2022-02-08T09:40:31.798686Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#ImageId,Label\n\npd.DataFrame({\"ImageId\": list(range(1,len(d)+1)), \"Label\": d}).to_csv('output.csv', index=False, header=True)","metadata":{"_uuid":"49dca1ec26ac9b6f3c7e5425a74d1f9472a66f61","_execution_state":"idle","_cell_guid":"9ee8814b-6ade-499e-94c5-050e85801e0b","execution":{"iopub.status.busy":"2022-02-08T09:40:31.800801Z","iopub.execute_input":"2022-02-08T09:40:31.801176Z","iopub.status.idle":"2022-02-08T09:40:31.877578Z","shell.execute_reply.started":"2022-02-08T09:40:31.801039Z","shell.execute_reply":"2022-02-08T09:40:31.876918Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## REFERENCES\n\nhttp://users.isr.ist.utl.pt/~wurmd/Livros/school/Bishop%20-%20Pattern%20Recognition%20And%20Machine%20Learning%20-%20Springer%20%202006.pdf","metadata":{"_uuid":"7ba2dc5e498f5913d188f20f41a29c2896d5963a","_execution_state":"idle","_cell_guid":"b4d15e48-bd8b-4471-9f7f-16770a9f7a28"}}]}