clases/notebooks/02 Perceptron.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Clasificación con el Algoritmo del Perceptrón\n",
    "\n",
    "Implementaremos una versión básica del algoritmo del perceptrón para clasificación binaria.\n",
    "\n",
    "En este caso los datos de entrada serán de **dos dimensiones**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "np.set_printoptions(suppress=True)  # no usar notacion \"e\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_decision_boundary(model, X, y):\n",
    "    X = X.T\n",
    "    y = y.T\n",
    "\n",
    "    # Set min and max values and give it some padding\n",
    "    x_min, x_max = X[0, :].min() - 1, X[0, :].max() + 1\n",
    "    y_min, y_max = X[1, :].min() - 1, X[1, :].max() + 1\n",
    "    h = 0.01\n",
    "\n",
    "    # Generate a grid of points with distance h between them\n",
    "    xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h))\n",
    "\n",
    "    # Predict the function value for the whole grid\n",
    "    Z = model(np.c_[xx.ravel(), yy.ravel()])\n",
    "    Z = Z.reshape(xx.shape)\n",
    "\n",
    "    ax = plt.gca()\n",
    "    ax.set_aspect(1)\n",
    "    # plt.axis(\"equal\")\n",
    "    #plt.contourf(xx, yy, Z, cmap=plt.cm.ocean, alpha=0.5)\n",
    "    plt.contourf(xx, yy, Z, colors = [\"red\",\"royalblue\"], alpha=0.2)\n",
    "    # plt.contourf(xx, yy, Z, cmap=plt.cm.Pastel1, alpha=0.5)\n",
    "    # plt.scatter(X[0, y==1], X[1, y==1], color=\"dodgerblue\", edgecolors='k', label=\"1\")\n",
    "    plt.scatter(X[0, y == 1], X[1, y == 1], color=\"royalblue\", label=\"1\")\n",
    "    plt.scatter(X[0, y == -1], X[1, y == -1], color=\"red\", label=\"-1\")\n",
    "    plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Datos\n",
    "\n",
    "Usaremos scikit-learn ([make_classification](https://scikit-learn.org/stable/modules/generated/sklearn.datasets.make_classification.html)) para generar datos aleatorios de dos dimensiones, divididos en dos clases etiquetadas con los valores 1 y -1.\n",
    "\n",
    "Los datos serán linealmente separables. Fijamos la semilla aleatoria para obtener siempre los mismos resultados."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import make_classification\n",
    "\n",
    "size = 100\n",
    "X, y_true = make_classification(n_samples=size, n_features=2, n_classes=2,\n",
    "                                   n_redundant=0, n_informative=2,\n",
    "                                   n_clusters_per_class=1, class_sep=0.5,\n",
    "                                   random_state=1)\n",
    "y_true[y_true == 0] = -1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100, 2)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(100,)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_true.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Graficamos para ver el resultado. Acá, los azules son 1, los rojos -1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X[y_true==1, 0], X[y_true==1, 1], color=\"royalblue\", label=\"1\")\n",
    "plt.scatter(X[y_true==-1, 0], X[y_true==-1, 1], color=\"red\", label=\"-1\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### División en Entrenamiento y Evaluación"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_size = 60\n",
    "val_size = size - train_size\n",
    "X_train, X_val = X[:train_size], X[train_size:]\n",
    "y_train, y_val = y_true[:train_size], y_true[train_size:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((60, 2), (40, 2))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape, X_val.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Algoritmo del Perceptrón"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inicialización\n",
    "\n",
    "Comenzamos a partir de un vector cualquiera:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "w =np.ones(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1., 1.])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predicción\n",
    "\n",
    "Definimos la función de predicción:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predict(X, w):\n",
    "    return np.sign(np.dot(X, w))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Probamos predecir con el primer modelo:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.,  1., -1.,  1.,  1., -1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,\n",
       "       -1., -1., -1.,  1.,  1., -1., -1.,  1.,  1.,  1., -1., -1., -1.,\n",
       "        1.,  1., -1., -1.,  1., -1.,  1.,  1., -1.,  1., -1., -1.,  1.,\n",
       "        1.])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = predict(X_val, w)\n",
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1, -1,  1, -1, -1,  1, -1, -1, -1,  1, -1, -1, -1,  1,  1,  1, -1,\n",
       "        1,  1,  1, -1, -1, -1,  1,  1,  1, -1,  1,  1,  1, -1,  1, -1, -1,\n",
       "        1, -1,  1,  1, -1,  1])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_val"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluación\n",
    "\n",
    "Calcularemos la \"accuracy\" sobre los datos de evaluación."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(y_true, y_pred):\n",
    "    return (y_true == y_pred).sum() / y_true.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accuracy(y_val, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "En este caso el resultado es muy bajo.\n",
    "\n",
    "Graficamos la frontera de decisión:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def predfun(x):\n",
    "    return predict(x, w)\n",
    "    \n",
    "plot_decision_boundary(predfun, \n",
    "                       X_val, \n",
    "                       y_val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Vemos que la región rosa debería contener a los puntos rojos y la celeste a los puntos azules."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Un Paso del Algoritmo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Primero debemos buscar un elemento de entrenamiento mal clasificado:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = predict(X_train, w)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1.,  1.,  1., -1.,  1.,  1.,  1., -1.,  1.,  1.,  1.,  1.,  1.,\n",
       "        1.,  1.,  1.,  1., -1.,  1., -1.,  1.,  1., -1.,  1.,  1., -1.,\n",
       "        1.,  1.,  1.,  1., -1., -1., -1.,  1.,  1., -1.,  1.,  1.,  1.,\n",
       "        1., -1., -1.,  1.,  1., -1.,  1., -1., -1.,  1., -1., -1.,  1.,\n",
       "        1.,  1., -1.,  1.,  1.,  1.,  1., -1.])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 0,  1,  2,  3,  6,  7,  8, 10, 11, 12, 13, 14, 15, 17, 19, 21, 22,\n",
       "        23, 24, 25, 26, 27, 28, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40,\n",
       "        41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 54, 55, 56, 57, 59],\n",
       "       dtype=int64),)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.where(y_train != y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, -1.0)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train[0], y_pred[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tomamos el primero:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "i = 0\n",
    "xi, yi = X_train[i], y_train[i]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2,)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xi.shape  # qué da?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Verificamos que la clasificación es incorrecta:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, -1.0)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yi_pred = predict(xi, w)\n",
    "yi, yi_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Probemos graficar el punto:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y2 = np.ones(size)\n",
    "y2[i] = -1\n",
    "plot_decision_boundary(lambda x: predict(x, w), X, y2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ahora actualicemos el modelo y veamos si se corrige la clasificación:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r = 1.0  # learning rate\n",
    "new_w = w + r * xi * yi\n",
    "predict(xi, new_w)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Probemos otras tasas de aprendizaje, a ver qué pasa:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.001 -1.0\n",
      "0.01 -1.0\n",
      "0.1 -1.0\n",
      "0.5 1.0\n",
      "1.0 1.0\n"
     ]
    }
   ],
   "source": [
    "for r in [0.001, 0.01, 0.1, 0.5, 1.0]:\n",
    "    new_w = w + r * xi * yi\n",
    "    print(r, predict(xi, new_w))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lo dejamos en $r=1$. Evaluemos el nuevo modelo, para ver si mejora el resultado:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = predict(X_val, new_w)\n",
    "accuracy(y_val, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_decision_boundary(lambda x: predict(x, new_w), X, y_true)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.45051362, 1.3786438 ])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_w"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Perceptrón: Algoritmo completo\n",
    "\n",
    "Implementaremos el algoritmo completo, iterando hasta que todos los datos de entrenamiento estén clasificados correctamente:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "step=0\ttrain_acc=0.17\tval_acc=0.1\n",
      "step=1\ttrain_acc=0.32\tval_acc=0.17\n",
      "step=2\ttrain_acc=0.37\tval_acc=0.28\n",
      "step=3\ttrain_acc=0.37\tval_acc=0.3\n",
      "step=4\ttrain_acc=0.48\tval_acc=0.33\n",
      "step=5\ttrain_acc=0.52\tval_acc=0.5\n",
      "step=6\ttrain_acc=0.78\tval_acc=0.8\n",
      "step=7\ttrain_acc=0.97\tval_acc=0.95\n",
      "step=8\ttrain_acc=0.87\tval_acc=0.85\n",
      "step=9\ttrain_acc=1.0\tval_acc=1.0\n"
     ]
    }
   ],
   "source": [
    "w = np.ones(2)\n",
    "\n",
    "r = 0.5  # qué pasaría\n",
    "step = 0\n",
    "finished = False\n",
    "while not finished:\n",
    "    # 1. predecir usando el modelo actual\n",
    "    y_pred = predict(X_train, w)\n",
    "    y_val_pred = predict(X_val, w)\n",
    "    \n",
    "    # 2. evaluar\n",
    "    train_acc = accuracy(y_train, y_pred)\n",
    "    val_acc = accuracy(y_val, y_val_pred)\n",
    "    print(f'step={step}\\ttrain_acc={train_acc:0.2}\\tval_acc={val_acc:0.2}')\n",
    "\n",
    "    # 3. si no clasifica todo bien, actualizar modelo\n",
    "    if (y_train == y_pred).all():\n",
    "        finished = True\n",
    "    else:\n",
    "        indices = np.where(y_train != y_pred)\n",
    "        i = indices[0][0]\n",
    "        xi, yi = X_train[i], y_train[i]\n",
    "        w = w + r * xi * yi\n",
    "        \n",
    "        step = step + 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Grafiquemos el resultado final:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_decision_boundary(lambda x: predict(x, w), X, y_true)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.67429978,  0.02427719])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ejercicios\n",
    "\n",
    "1. Experimentar con otras tasas de aprendizaje.\n",
    "2. Generar datos no linealmente separables (probar achicando el valor de 'class_sep' en 'make_classification').\n",
    "3. Adaptar el algoritmo del perceptrón para que funcione con estos datos (ver variante \"pocket\" en [Wikipedia](https://en.wikipedia.org/wiki/Perceptron)).\n",
    "4. Hacer todo con el perceptrón de Scikit-learn ([Perceptron](https://scikit-learn.org/stable/modules/linear_model.html#perceptron))."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Referencias\n",
    "\n",
    "- [Perceptron (Wikipedia)](https://en.wikipedia.org/wiki/Perceptron)\n",
    "- [Perceptron Learning Algorithm: A Graphical Explanation Of Why It Works](https://towardsdatascience.com/perceptron-learning-algorithm-d5db0deab975)\n",
    "- Gallant, S. I. (1990). [Perceptron-based learning algorithms.](https://ftp.cs.nyu.edu/~roweis/csc2515-2006/readings/gallant.pdf) IEEE Transactions on Neural Networks, vol. 1, no. 2, pp. 179–191. \n",
    "\n",
    "Scikit-learn:\n",
    "\n",
    "- [make_classification](https://scikit-learn.org/stable/modules/generated/sklearn.datasets.make_classification.html)\n",
    "- [Perceptron](https://scikit-learn.org/stable/modules/linear_model.html#perceptron)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}