digit recognizer.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0444e160",
   "metadata": {},
   "source": [
    "The goal of this project is to build a multilayer perceptron multiclass classifier from scratch using only numpy. The neural net that I created is a proof of concept and a display of my understanding of how MLPs work at the low level. \n",
    "\n",
    "Here is a quick layout of the notebook:\n",
    "\n",
    "- I first take the dataset from the [Digit Recognizer Kaggle challenge](https://www.kaggle.com/competitions/digit-recognizer) and use the neural network I created to make a submission and get an evaluation\n",
    "- Then I create a network of the same architecture and API using PyTorch\n",
    "- After that I compare the results and discuss the shortcomings of my numpy neural net\n",
    "\n",
    "This notebook uses my networks as modules, the source code for which can be found in the corresponding folders. "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "eb5c737c",
   "metadata": {},
   "source": [
    "# Part 1 - numpy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "59e076ea-ef45-4057-9977-943ea6e4edac",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "from my_numpy_nn.mlp_classifier import MLPClassifier"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4f25120c",
   "metadata": {},
   "source": [
    "Here is the description of the MLPClassifier and its init parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "7e0834b8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    MLP multiclass classifier with an arbitrary number\n",
      "     of fully connected linear layers and softmax cross-entropy loss function\n",
      "    \n",
      "\n",
      "        Arguments:\n",
      "            input_size: (int): size of 1 instance of data (X)\n",
      "            \n",
      "            output_size: (int): size of the prediction vector\n",
      "            \n",
      "            n_layers: (int): number of layers in the network\n",
      "            \n",
      "            layer_dims: list[int]: dimensions of hidden layers,\n",
      "            (len(layer_dims) = n_layers - 1)\n",
      "        \n"
     ]
    }
   ],
   "source": [
    "print(MLPClassifier.__doc__)\n",
    "print(MLPClassifier.__init__.__doc__)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "ecdb954e",
   "metadata": {},
   "source": [
    "Import the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2085c288-49c2-4a03-9cd9-17c1dcafc5f2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>pixel0</th>\n",
       "      <th>pixel1</th>\n",
       "      <th>pixel2</th>\n",
       "      <th>pixel3</th>\n",
       "      <th>pixel4</th>\n",
       "      <th>pixel5</th>\n",
       "      <th>pixel6</th>\n",
       "      <th>pixel7</th>\n",
       "      <th>pixel8</th>\n",
       "      <th>...</th>\n",
       "      <th>pixel774</th>\n",
       "      <th>pixel775</th>\n",
       "      <th>pixel776</th>\n",
       "      <th>pixel777</th>\n",
       "      <th>pixel778</th>\n",
       "      <th>pixel779</th>\n",
       "      <th>pixel780</th>\n",
       "      <th>pixel781</th>\n",
       "      <th>pixel782</th>\n",
       "      <th>pixel783</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 785 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   label  pixel0  pixel1  pixel2  pixel3  pixel4  pixel5  pixel6  pixel7   \n",
       "0      1       0       0       0       0       0       0       0       0  \\\n",
       "1      0       0       0       0       0       0       0       0       0   \n",
       "2      1       0       0       0       0       0       0       0       0   \n",
       "3      4       0       0       0       0       0       0       0       0   \n",
       "4      0       0       0       0       0       0       0       0       0   \n",
       "\n",
       "   pixel8  ...  pixel774  pixel775  pixel776  pixel777  pixel778  pixel779   \n",
       "0       0  ...         0         0         0         0         0         0  \\\n",
       "1       0  ...         0         0         0         0         0         0   \n",
       "2       0  ...         0         0         0         0         0         0   \n",
       "3       0  ...         0         0         0         0         0         0   \n",
       "4       0  ...         0         0         0         0         0         0   \n",
       "\n",
       "   pixel780  pixel781  pixel782  pixel783  \n",
       "0         0         0         0         0  \n",
       "1         0         0         0         0  \n",
       "2         0         0         0         0  \n",
       "3         0         0         0         0  \n",
       "4         0         0         0         0  \n",
       "\n",
       "[5 rows x 785 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "digits = pd.read_csv('train.csv')\n",
    "digits.head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "b6ff6075",
   "metadata": {},
   "source": [
    "Detach labels and normalize X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "cd9cd0bd-7c8c-44de-9dd7-6e93ddab8a02",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = digits['label']\n",
    "X_train = digits.drop(['label'], axis=1) / 255 "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "bfe8202b",
   "metadata": {},
   "source": [
    "Print the first 5 images with their labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "12060696-b024-4555-9739-e338c96b8c91",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_samples = 5\n",
    "f, ax  = plt.subplots(1,n_samples, figsize=(n_samples, 5))\n",
    "\n",
    "for i in range(n_samples):\n",
    "    ax[i].imshow(X_train.loc[i, :].values.reshape(28,28))\n",
    "    ax[i].axis('off')\n",
    "    ax[i].set_title('target: '+str(y_train[i]))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "f4b98d12",
   "metadata": {},
   "source": [
    "Set hyperparameteres"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "eb5f93f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "input_size = X_train.shape[1]\n",
    "output_size = len(set(y_train))\n",
    "layer_dims = [100, 350, 500]\n",
    "n_layers = len(layer_dims) + 1\n",
    "n_epochs = 20\n",
    "learning_rate = 1e-3"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "d9f3e0a2",
   "metadata": {},
   "source": [
    "Create a classifier and train it, measure accuracy on the same data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c3c3e5e8-283b-4f40-902a-8f01bc2919f9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 1/20\tloss: 0.327\t balanced accuracy on train: 0.899\n",
      "Epoch: 2/20\tloss: 0.153\t balanced accuracy on train: 0.954\n",
      "Epoch: 3/20\tloss: 0.108\t balanced accuracy on train: 0.967\n",
      "Epoch: 4/20\tloss: 0.083\t balanced accuracy on train: 0.976\n",
      "Epoch: 5/20\tloss: 0.065\t balanced accuracy on train: 0.981\n",
      "Epoch: 6/20\tloss: 0.052\t balanced accuracy on train: 0.986\n",
      "Epoch: 7/20\tloss: 0.042\t balanced accuracy on train: 0.989\n",
      "Epoch: 8/20\tloss: 0.034\t balanced accuracy on train: 0.992\n",
      "Epoch: 9/20\tloss: 0.027\t balanced accuracy on train: 0.994\n",
      "Epoch: 10/20\tloss: 0.022\t balanced accuracy on train: 0.996\n",
      "Epoch: 11/20\tloss: 0.017\t balanced accuracy on train: 0.997\n",
      "Epoch: 12/20\tloss: 0.014\t balanced accuracy on train: 0.998\n",
      "Epoch: 13/20\tloss: 0.011\t balanced accuracy on train: 0.999\n",
      "Epoch: 14/20\tloss: 0.009\t balanced accuracy on train: 0.999\n",
      "Epoch: 15/20\tloss: 0.007\t balanced accuracy on train: 1.000\n",
      "Epoch: 16/20\tloss: 0.006\t balanced accuracy on train: 1.000\n",
      "Epoch: 17/20\tloss: 0.005\t balanced accuracy on train: 1.000\n",
      "Epoch: 18/20\tloss: 0.004\t balanced accuracy on train: 1.000\n",
      "Epoch: 19/20\tloss: 0.003\t balanced accuracy on train: 1.000\n",
      "Epoch: 20/20\tloss: 0.003\t balanced accuracy on train: 1.000\n"
     ]
    }
   ],
   "source": [
    "mlp = MLPClassifier(input_size=input_size, output_size=output_size, \n",
    "                    n_layers=n_layers, layer_dims=layer_dims)\n",
    "mlp.fit(X_train.values, y_train, n_epochs=n_epochs, \n",
    "        learning_rate=learning_rate)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8af7adc5",
   "metadata": {},
   "source": [
    "Plot metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "42ad558a-c148-45e1-933e-2a7d5832ae64",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1750 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils.plot_metrics import PlotMetrics\n",
    "PlotMetrics(mlp).plot()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "c2ed38a1",
   "metadata": {},
   "source": [
    "Make and save a prediction on the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "4b90f839-93bb-4cdc-a663-2cb0eef18dbe",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = pd.read_csv('test.csv')\n",
    "y_pred = mlp.predict(X_test.values)\n",
    "y_pred = pd.DataFrame(y_pred, columns= ['Label'])\n",
    "y_pred.index += 1\n",
    "y_pred['Label'].to_csv('submission.csv', index_label='ImageId')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69f329d0-afe6-430b-af00-934742e49972",
   "metadata": {},
   "source": [
    "Resulting accuracy on contest data: 0.97132"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d38751a9-f892-4ab9-92af-a819c4cb3d2d",
   "metadata": {},
   "source": [
    "![result1](https://raw.githubusercontent.com/Shmuley322/digit-recognizer/master/results/result1.PNG)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "e1a67cd5",
   "metadata": {},
   "source": [
    "# Part 2 - PyTorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "9a892b5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "\n",
    "from my_torch_nn.mlp_classifier import MLPClassifierTorch\n",
    "from my_torch_nn.numbers_dataset import NumbersDataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a92c990e-cc2f-49e9-ae4c-bcd4edbe73d3",
   "metadata": {},
   "source": [
    "Create a torch Dataset from train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "a8967bb3",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = NumbersDataset('train.csv')\n",
    "train_loader = DataLoader(dataset=train_data, batch_size=64, shuffle=True, num_workers=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02ed0d20-66bc-44cf-9197-a22bc231764e",
   "metadata": {},
   "source": [
    "Initialize a classifier and train it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "68a5cb6f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 1/20\tloss: 0.310\t balanced accuracy on train: 0.959\n",
      "Epoch: 2/20\tloss: 0.151\t balanced accuracy on train: 0.960\n",
      "Epoch: 3/20\tloss: 0.120\t balanced accuracy on train: 0.974\n",
      "Epoch: 4/20\tloss: 0.106\t balanced accuracy on train: 0.979\n",
      "Epoch: 5/20\tloss: 0.095\t balanced accuracy on train: 0.978\n",
      "Epoch: 6/20\tloss: 0.087\t balanced accuracy on train: 0.980\n",
      "Epoch: 7/20\tloss: 0.085\t balanced accuracy on train: 0.971\n",
      "Epoch: 8/20\tloss: 0.077\t balanced accuracy on train: 0.986\n",
      "Epoch: 9/20\tloss: 0.070\t balanced accuracy on train: 0.984\n",
      "Epoch: 10/20\tloss: 0.065\t balanced accuracy on train: 0.986\n",
      "Epoch: 11/20\tloss: 0.061\t balanced accuracy on train: 0.987\n",
      "Epoch: 12/20\tloss: 0.066\t balanced accuracy on train: 0.989\n",
      "Epoch: 13/20\tloss: 0.061\t balanced accuracy on train: 0.990\n",
      "Epoch: 14/20\tloss: 0.054\t balanced accuracy on train: 0.990\n",
      "Epoch: 15/20\tloss: 0.056\t balanced accuracy on train: 0.990\n",
      "Epoch: 16/20\tloss: 0.050\t balanced accuracy on train: 0.987\n",
      "Epoch: 17/20\tloss: 0.047\t balanced accuracy on train: 0.988\n",
      "Epoch: 18/20\tloss: 0.052\t balanced accuracy on train: 0.986\n",
      "Epoch: 19/20\tloss: 0.052\t balanced accuracy on train: 0.991\n",
      "Epoch: 20/20\tloss: 0.042\t balanced accuracy on train: 0.992\n"
     ]
    }
   ],
   "source": [
    "mlp_torch = MLPClassifierTorch(input_size=input_size, output_size=output_size, \n",
    "                    n_layers=n_layers, layer_dims=layer_dims)\n",
    "mlp_torch.fit(train_loader, learning_rate=learning_rate, n_epochs=n_epochs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de3f2362-854d-42b9-9c38-4a4bb3160177",
   "metadata": {},
   "source": [
    "Plot metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ead50e7a-9563-41b6-bd45-d3e88590555c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1400 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "PlotMetrics(mlp_torch).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b545d885-efd2-465c-8af9-334f0f98f96f",
   "metadata": {},
   "source": [
    "Make and save a prediction on the test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "297cfac8",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = pd.read_csv('test.csv')\n",
    "y_pred = mlp_torch.predict(torch.tensor(X_test.values, dtype=torch.float32))\n",
    "y_pred = pd.DataFrame(y_pred, columns= ['Label'])\n",
    "y_pred.index += 1\n",
    "y_pred['Label'].to_csv('submission2.csv', index_label='ImageId')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c219a0a6-dc55-4c7e-9b3d-0051c43a7f7f",
   "metadata": {},
   "source": [
    "Resulting accuracy on contest data: 0.96671"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0c0ed98-c4f4-4fc3-9802-87f5f884a3a1",
   "metadata": {},
   "source": [
    "![result2](https://raw.githubusercontent.com/Shmuley322/digit-recognizer/master/results/result2.PNG)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8f0104b-c7f8-4c1b-b234-decdd7434660",
   "metadata": {},
   "source": [
    "# Part 3 - discussion"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "817ddab0-3fce-46b3-a321-6a27eb49d177",
   "metadata": {},
   "source": [
    "Both models performed very similarly, however, the numpy neural net has 2 major issues:\n",
    "1. Some choices of layer dimensions may lead to overflow in the activation function during forward pass and oppositely vanishing gradients during backpropagation\n",
    "2. All computations take place on the CPU which makes the training process really slow\n",
    "\n",
    "It is my understanding that the core of the first problem lies in poor initialization of weights and biases for the layers. I am using samples from the [standard normal distribution](https://numpy.org/doc/stable/reference/random/generated/numpy.random.randn.html) of mean 0 and variance 1 further modified by dividing by 10. Instead of dividing by 10 I could make use of Xavier and He initializers as described [here](https://github.com/christianversloot/machine-learning-articles/blob/main/he-xavier-initialization-activation-functions-choose-wisely.md).\n",
    "\n",
    "To solve the second problem I could use libraries like gnumpy or CuPy which would help accelerate the computations.\n",
    "\n",
    "Implementing these things would majorly improve training time and robustness of the network, but at this point I am satisfied with what I have achieved and have decided to move on to other projects. "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "FastKernel",
   "language": "python",
   "name": "fastkernel"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}