SNN_vs_BatchNorm-MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Comparision of convergence speed and accuracy of SNNs over standard FNNs with Batch Normalisation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing : MNIST dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from tqdm import tqdm\n",
    "import torch.nn as nn\n",
    "import matplotlib.pyplot as plt\n",
    "from torch.utils.data import DataLoader\n",
    "from keras.datasets.mnist import load_data\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "sns.set_theme()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load (and normalize) mnist dataset\n",
    "(trainX, trainy), (testX, testy) = load_data()\n",
    "s = StandardScaler()\n",
    "s.fit(trainX.reshape(-1, 28 * 28))\n",
    "trainX = s.transform(trainX.reshape(-1, 28 * 28)).astype(np.float32)\n",
    "testX = s.transform(testX.reshape(-1, 28 * 28)).astype(np.float32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Defining models : SNN and MLP( with BatchNorm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class SELU(nn.Module):\n",
    "\n",
    "    def __init__(self):\n",
    "        super(SELU, self).__init__()\n",
    "        # the parameters alpha and lambda_ are obtained in getSELUparameters.ipynb\n",
    "        self.alpha = 1.6732632423543772848170429916717\n",
    "        self.lambda_ = 1.0507009873554804934193349852946\n",
    "\n",
    "    def forward(self, x):\n",
    "        # SELU activation function\n",
    "        # piecewise activation function\n",
    "        # x < 0: lambda_ * alpha * (exp(x) - 1)\n",
    "        # x >= 0: lambda_ * x\n",
    "        return self.lambda_ * (torch.maximum(torch.tensor([0], device=x.device), x) + \n",
    "                               torch.minimum(torch.tensor([0], device=x.device),self.alpha * (torch.exp(x) - 1))\n",
    "                               )\n",
    "\n",
    "\n",
    "class SNN(nn.Module):\n",
    "\n",
    "    def __init__(self, input_dim=28 * 28, hidden_dim=28 * 28, output_dim=10, depth=8):\n",
    "        super(SNN, self).__init__()\n",
    "\n",
    "        model = []\n",
    "        # add layers equal to the depth of the network\n",
    "        for _ in range(depth):\n",
    "            # each layer is a linear layer followed by a SELU activation\n",
    "            model += [nn.Linear(input_dim, hidden_dim), SELU()]\n",
    "        \n",
    "        # add the final layer as a linear layer followed by a log softmax\n",
    "        # this is to ensure that the output is a probability distribution\n",
    "        model += [nn.Linear(hidden_dim, output_dim), nn.LogSoftmax(dim=-1)]\n",
    "        self.network = nn.Sequential(*model)\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.network(x)\n",
    "\n",
    "\n",
    "class MLP(nn.Module):\n",
    "\n",
    "    def __init__(self, input_dim=28 * 28, hidden_dim=28 * 28, output_dim=10, depth=8):\n",
    "        super(MLP, self).__init__()\n",
    "\n",
    "        model = []\n",
    "        # add layers equal to the depth of the network\n",
    "        for _ in range(depth):\n",
    "            # each layer is a linear layer followed by a Batch Normalisation layer\n",
    "            model += [nn.Linear(input_dim, hidden_dim), nn.ReLU(), torch.nn.BatchNorm1d(hidden_dim)]\n",
    "\n",
    "        # add the final layer as a linear layer followed by a log softmax\n",
    "        # this is to ensure that the output is a probability distribution\n",
    "        model += [nn.Linear(hidden_dim, output_dim), nn.LogSoftmax(dim=-1)]\n",
    "        self.network = nn.Sequential(*model)\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.network(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Weights initialisation and train loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(model, optimizer, dataset, loss_fct=torch.nn.NLLLoss(), device='cpu', nb_epochs=25):\n",
    "    training_loss = []\n",
    "\n",
    "    # Loop over epochs\n",
    "    for _ in tqdm(range(nb_epochs)):\n",
    "        batch_loss = []\n",
    "        # Loop over batches in the dataset\n",
    "        for batch in dataset:\n",
    "            # Extract features (x) and labels (y) from the batch\n",
    "            x, y = batch\n",
    "\n",
    "            # Forward pass: compute predicted log probabilities\n",
    "            log_prob = model(x.to(device).reshape(-1, 28 * 28))\n",
    "\n",
    "            # Calculate the loss between predicted log probabilities and actual labels\n",
    "            loss = loss_fct(log_prob, y.to(device))\n",
    "\n",
    "            # Zero the gradients to avoid accumulation\n",
    "            optimizer.zero_grad()\n",
    "\n",
    "            # Backward pass\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "\n",
    "            # Append the current batch's loss to the list\n",
    "            batch_loss.append(loss.item())\n",
    "\n",
    "        # Compute the mean training loss for the current epoch\n",
    "        epoch_loss = np.mean(batch_loss)\n",
    "\n",
    "        # Append the mean training loss to the list for all epochs\n",
    "        training_loss.append(epoch_loss)\n",
    "\n",
    "    # Return the list of training losses for each epoch\n",
    "    return training_loss\n",
    "\n",
    "def _init_weights(module, init_gain=1.):\n",
    "    # initialise weights using lecun normal form for selu\n",
    "    # initialise weights using MSRA initialisation for MLP batch normalisation\n",
    "    if isinstance(module, nn.Linear):\n",
    "        # lecun Normal :    mean=0 and var=1/n where n is the number of input features\n",
    "        # MSRA         :    mean=0 and var=2/n where n is the number of input features\n",
    "        module.weight.data.normal_(mean=0.0, std=np.sqrt(init_gain / module.weight.data.shape[0]))\n",
    "        if module.bias is not None:\n",
    "            module.bias.data.zero_()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = 'cuda'\n",
    "nb_epochs = 2000\n",
    "dataset = DataLoader([[trainX[i], trainy[i]] for i in range(trainX.shape[0])], batch_size=128, shuffle=True)\n",
    "\n",
    "# consider the depth of the network as a hyperparameter\n",
    "depth_parameters = [8, 16, 32]\n",
    "\n",
    "# train the snn models with varying depths and plot the training error\n",
    "training_loss_snn = {}\n",
    "for depth in depth_parameters:\n",
    "    model = SNN(depth=depth).to(device)\n",
    "    # intialise weights using lecun normal form\n",
    "    model.apply(_init_weights)\n",
    "    optimizer = torch.optim.SGD(model.parameters(), lr=1e-5)\n",
    "    training_loss_snn[depth] = train(model, optimizer, dataset, device=device, nb_epochs=nb_epochs)\n",
    "\n",
    "# train the mlp models with varying depths and plot the training error\n",
    "training_loss_mlp = {}\n",
    "for depth in depth_parameters:\n",
    "    model = MLP(depth=depth).to(device)\n",
    "    # intialise weights using MSRA initialisation\n",
    "    model.apply(lambda x: _init_weights(x, init_gain=2.))\n",
    "    optimizer = torch.optim.SGD(model.parameters(), lr=1e-5)\n",
    "    training_loss_mlp[depth] = train(model, optimizer, dataset, device=device, nb_epochs=nb_epochs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plotting Epochs vs Loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot():\n",
    "    # plot the training error and save the image\n",
    "    fontsize = 14\n",
    "    plt.plot(training_loss_mlp[8], color=plt.cm.bwr([0.05]), label='BatchNorm Depth 8')\n",
    "    plt.plot(training_loss_mlp[16], color=plt.cm.bwr([0.15]), label='BatchNorm Depth 16')\n",
    "    plt.plot(training_loss_mlp[32], color=plt.cm.bwr([0.25]), label='BatchNorm Depth 32')\n",
    "    plt.plot(training_loss_snn[8], color=plt.cm.bwr([0.75]), label='SNN Depth 8')\n",
    "    plt.plot(training_loss_snn[16], color=plt.cm.bwr([0.85]), label='SNN Depth 16')\n",
    "    plt.plot(training_loss_snn[32], color=plt.cm.bwr([0.95]), label='SNN Depth 32')\n",
    "    plt.yscale(\"log\")\n",
    "    plt.xlabel(\"Iterations\", fontsize=fontsize)\n",
    "    plt.ylabel(\"Training Error\", fontsize=fontsize)\n",
    "    plt.legend()\n",
    "    plt.savefig('snn_vs_mlp.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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",
      "image/png": "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",
      "text/plain": [
       "<PIL.PngImagePlugin.PngImageFile image mode=RGB size=398x266>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# display the final image\n",
    "from IPython.display import display\n",
    "from PIL import Image as PILImage\n",
    "\n",
    "image_path = './snn_vs_mlp.png'\n",
    "\n",
    "# Open the image\n",
    "original_image = PILImage.open(image_path)\n",
    "\n",
    "# Display the image with the white background\n",
    "display(original_image)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### FNNs with batch normalization exhibit high variance due to perturbations. In contrast, SNNs do not suffer from high variance as they are more robust to perturbations and learn faster"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}