PatiñoAngel_RedNeuronal (1).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "**EJERCICIO UNO CON UNA SOLA CAPA POR ANGEL PATIÑO**"
      ],
      "metadata": {
        "id": "CPqxuHI4H08R"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hy_pAhzDZ4gy"
      },
      "source": [
        "#importamos las librerias\n",
        "import tensorflow as tf # libreria de IA\n",
        "import numpy as np #libreria para manejar arreglos\n",
        "import matplotlib.pyplot as plt #graficar"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MIrfX72vaL1z"
      },
      "source": [
        "#Ejemplos para aprender\n",
        "celsius = np.array([-40, -10, 0, 8, 15, 22, 38], dtype=float) #entrada de grados celsius\n",
        "fahrenheit = np.array([-40, 14, 32, 46, 59, 72, 100], dtype=float) #entrada de grados celsius"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RIG6Y5jEafFL"
      },
      "source": [
        "capa = tf.keras.layers.Dense(units=1, input_shape=[1]) # UNA NEURONA DENSA\n",
        "modelo = tf.keras.Sequential([capa]) #MODELO SECUENCIAL"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1_WHzWidbN_8"
      },
      "source": [
        "modelo.compile(\n",
        "    optimizer=tf.keras.optimizers.Adam(0.1), #Optimizador ADAM (Ajusta sesgo y peso automaticamente eficientemente, el parametro es la tasa de aprendizaje)\n",
        "    loss='mean_squared_error' #error cuadrado medio\n",
        ")"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4D-NuUTnbabR",
        "outputId": "cdb75818-2e92-4770-c442-4c49bcad34d9"
      },
      "source": [
        "print(\"Comenzando entrenamiento...\")\n",
        "historial = modelo.fit(celsius, fahrenheit, epochs=1000, verbose=False) #fit entrena, pasamos los datos de celsius y fahrenheit, y las vueltas\n",
        "print(\"Modelo entrenado!\")"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Comenzando entrenamiento...\n",
            "Modelo entrenado!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "px43cF3Nb3mG",
        "outputId": "54fde9cd-4db0-4c38-fe40-0c1c5c6c8959"
      },
      "source": [
        "# funcion de perdidad\n",
        "plt.xlabel(\"# Epoca\")\n",
        "plt.ylabel(\"Magnitud de pérdida\")\n",
        "plt.plot(historial.history[\"loss\"])"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7ed8ad892410>]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "esVvhzjMcV9f",
        "outputId": "b13e0662-e03a-4b16-af74-5dbbb6de23f0"
      },
      "source": [
        "print(\"Prediccion\")\n",
        "resultado = modelo.predict([100.0])\n",
        "print(\"El resultado es: \" + str(resultado) + \" fahrenheit!\")"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Prediccion\n",
            "1/1 [==============================] - 0s 180ms/step\n",
            "El resultado es: [[211.74113]] fahrenheit!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FVDejrBgcokc",
        "outputId": "ba849530-91bc-45fb-a7f9-488ef4a036de"
      },
      "source": [
        "print(\"Variables internas del modelo\")\n",
        "print(capa.get_weights()) #obtener el peso, sesgo"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Variables internas del modelo\n",
            "[array([[1.798352]], dtype=float32), array([31.905926], dtype=float32)]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**SEGUNDO EJERCICIO DONDE ENTRENAMOS CON 3 CAPAS**"
      ],
      "metadata": {
        "id": "Nt3zEj1LH_-R"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# AUMENTAR CAPAS\n",
        "oculta1 = tf.keras.layers.Dense(units=3, input_shape=[1]) # PRIMERA CAPA\n",
        "oculta2 = tf.keras.layers.Dense(units=3) #SEGUNDA CAPA\n",
        "salida = tf.keras.layers.Dense(units=1) # CAPA DE SALIDA\n",
        "modelo = tf.keras.Sequential([oculta1, oculta2, salida]) # MODELO SECUENCIAL"
      ],
      "metadata": {
        "id": "kwZ3d-14IQ2r"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "modelo.compile(\n",
        "    optimizer=tf.keras.optimizers.Adam(0.1), #Optimizador ADAM (Ajusta sesgo y peso automaticamente eficientemente, el parametro es la tasa de aprendizaje)\n",
        "    loss='mean_squared_error' #error cuadrado medio\n",
        ")"
      ],
      "metadata": {
        "id": "JMK8YOwUJJPe"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"Comenzando entrenamiento...\")\n",
        "historial = modelo.fit(celsius, fahrenheit, epochs=1000, verbose=False) #fit entrena, pasamos los datos de celsius y fahrenheit, y las vueltas\n",
        "print(\"Modelo entrenado!\")"
      ],
      "metadata": {
        "id": "5GxBVfk2JMYS",
        "outputId": "2af62b9f-3d67-47dc-e95b-256a92839a56",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Comenzando entrenamiento...\n",
            "Modelo entrenado!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# funcion de perdidad\n",
        "plt.xlabel(\"# Epoca\")\n",
        "plt.ylabel(\"Magnitud de pérdida\")\n",
        "plt.title(\"Mayor capas\")\n",
        "plt.plot(historial.history[\"loss\"])"
      ],
      "metadata": {
        "id": "8PqIIxTnJO6Q",
        "outputId": "cdf9ebdd-b83c-42fa-f45e-c03e73164635",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        }
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7ed8ae82fac0>]"
            ]
          },
          "metadata": {},
          "execution_count": 12
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Si comparamos el modelo entranado de una capa (Ejercio Uno), con el modelo con tres capaz, observamos que obtenemos mejores resultados con este ultimo."
      ],
      "metadata": {
        "id": "fJ-tEjG6JpVJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# imprimir el sesgo y peso\n",
        "print(oculta1.get_weights())\n",
        "print(oculta2.get_weights())\n",
        "print(salida.get_weights())"
      ],
      "metadata": {
        "id": "R7UFI_nqItTx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3ab16fd0-722c-4ce8-bf8e-f45d22a9b4e3"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[array([[-0.05232973,  0.6475969 ,  0.21419732]], dtype=float32), array([2.4080932, 3.6836836, 3.4332693], dtype=float32)]\n",
            "[array([[ 0.59000593,  0.30340979, -0.25689414],\n",
            "       [ 0.9909865 , -0.12906489, -0.6849438 ],\n",
            "       [ 1.3118681 , -0.18349655,  0.5865532 ]], dtype=float32), array([ 3.7126238, -3.1693513, -3.6881018], dtype=float32)]\n",
            "[array([[ 1.4149656],\n",
            "       [-0.7831404],\n",
            "       [-1.4033202]], dtype=float32), array([3.6166015], dtype=float32)]\n"
          ]
        }
      ]
    }
  ]
}