concrete-strength-prediction-NN.ipynb

{
    "cells": [
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "# Concrete Strength Prediction Using Neural Networks in Keras "
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## Import Main Libraries "
        },
        {
            "cell_type": "code",
            "execution_count": 1,
            "metadata": {},
            "outputs": [],
            "source": "# main library imports \nimport pandas as pd \nimport numpy as np \nimport seaborn as sns \nimport matplotlib.pyplot as plt \nfrom scipy.stats import zscore \nfrom sklearn.metrics import mean_squared_error as MSE\nfrom sklearn.model_selection import train_test_split\nfrom math import sqrt\nfrom statistics import mean, stdev\nfrom IPython.display import clear_output\nimport time"
        },
        {
            "cell_type": "code",
            "execution_count": 2,
            "metadata": {},
            "outputs": [],
            "source": "# dowload keras and tensorflow and import\n# download and import keras to build the neural network \n#!pip install tensorFlow\n#!pip install keras\nfrom tensorflow import keras\nfrom tensorflow.keras.models import Sequential \nfrom tensorflow.keras.layers import Dense "
        },
        {
            "cell_type": "code",
            "execution_count": 3,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>Strength</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1040.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>79.99</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1055.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>61.89</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>332.5</td>\n      <td>142.5</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>594.0</td>\n      <td>270</td>\n      <td>40.27</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>332.5</td>\n      <td>142.5</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>594.0</td>\n      <td>365</td>\n      <td>41.05</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>198.6</td>\n      <td>132.4</td>\n      <td>0.0</td>\n      <td>192.0</td>\n      <td>0.0</td>\n      <td>978.4</td>\n      <td>825.5</td>\n      <td>360</td>\n      <td>44.30</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "   Cement  Blast Furnace Slag  Fly Ash  Water  Superplasticizer  \\\n0   540.0                 0.0      0.0  162.0               2.5   \n1   540.0                 0.0      0.0  162.0               2.5   \n2   332.5               142.5      0.0  228.0               0.0   \n3   332.5               142.5      0.0  228.0               0.0   \n4   198.6               132.4      0.0  192.0               0.0   \n\n   Coarse Aggregate  Fine Aggregate  Age  Strength  \n0            1040.0           676.0   28     79.99  \n1            1055.0           676.0   28     61.89  \n2             932.0           594.0  270     40.27  \n3             932.0           594.0  365     41.05  \n4             978.4           825.5  360     44.30  "
                    },
                    "execution_count": 3,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "concrete_data = pd.read_csv('https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DL0101EN/labs/data/concrete_data.csv')\nconcrete_data.head()"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Units: kg/m^3 for volumetric elements, MPa for Strength, days for age\n\n- Coming from a structural engineering background, I want to add another feature to this \ndataset. The feauture is the ratio of water to cement and it is often used by engineers to estimate concrete strength. I'll call this feature 'WC Ratio'"
        },
        {
            "cell_type": "code",
            "execution_count": 4,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>Strength</th>\n      <th>WC Ratio</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1040.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>79.99</td>\n      <td>0.300</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1055.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>61.89</td>\n      <td>0.300</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>332.5</td>\n      <td>142.5</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>594.0</td>\n      <td>270</td>\n      <td>40.27</td>\n      <td>0.686</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>332.5</td>\n      <td>142.5</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>594.0</td>\n      <td>365</td>\n      <td>41.05</td>\n      <td>0.686</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>198.6</td>\n      <td>132.4</td>\n      <td>0.0</td>\n      <td>192.0</td>\n      <td>0.0</td>\n      <td>978.4</td>\n      <td>825.5</td>\n      <td>360</td>\n      <td>44.30</td>\n      <td>0.967</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "   Cement  Blast Furnace Slag  Fly Ash  Water  Superplasticizer  \\\n0   540.0                 0.0      0.0  162.0               2.5   \n1   540.0                 0.0      0.0  162.0               2.5   \n2   332.5               142.5      0.0  228.0               0.0   \n3   332.5               142.5      0.0  228.0               0.0   \n4   198.6               132.4      0.0  192.0               0.0   \n\n   Coarse Aggregate  Fine Aggregate  Age  Strength  WC Ratio  \n0            1040.0           676.0   28     79.99     0.300  \n1            1055.0           676.0   28     61.89     0.300  \n2             932.0           594.0  270     40.27     0.686  \n3             932.0           594.0  365     41.05     0.686  \n4             978.4           825.5  360     44.30     0.967  "
                    },
                    "execution_count": 4,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "# add the new feature to the dataframe \nconcrete_data['WC Ratio'] = round(concrete_data['Water']/concrete_data['Cement'],3)\nconcrete_data.head()"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## Explore the Dataset "
        },
        {
            "cell_type": "code",
            "execution_count": 5,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "The dataframe has 1030 Rows and 10 Columns\n"
                }
            ],
            "source": "# check how many rows we have \nog_num_rows = concrete_data.shape[0]\nog_num_cols = concrete_data.shape[1]\nprint(f'The dataframe has {og_num_rows} Rows and {og_num_cols} Columns')"
        },
        {
            "cell_type": "code",
            "execution_count": 6,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": "Cement                False\nBlast Furnace Slag    False\nFly Ash               False\nWater                 False\nSuperplasticizer      False\nCoarse Aggregate      False\nFine Aggregate        False\nAge                   False\nStrength              False\nWC Ratio              False\ndtype: bool"
                    },
                    "execution_count": 6,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "# check if there is any missing data\nconcrete_data.isnull().any()"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- There is no data missing from any of the columns which is good "
        },
        {
            "cell_type": "code",
            "execution_count": 7,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>Strength</th>\n      <th>WC Ratio</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n      <td>1030.000000</td>\n    </tr>\n    <tr>\n      <th>mean</th>\n      <td>281.167864</td>\n      <td>73.895825</td>\n      <td>54.188350</td>\n      <td>181.567282</td>\n      <td>6.204660</td>\n      <td>972.918932</td>\n      <td>773.580485</td>\n      <td>45.662136</td>\n      <td>35.817961</td>\n      <td>0.748284</td>\n    </tr>\n    <tr>\n      <th>std</th>\n      <td>104.506364</td>\n      <td>86.279342</td>\n      <td>63.997004</td>\n      <td>21.354219</td>\n      <td>5.973841</td>\n      <td>77.753954</td>\n      <td>80.175980</td>\n      <td>63.169912</td>\n      <td>16.705742</td>\n      <td>0.314000</td>\n    </tr>\n    <tr>\n      <th>min</th>\n      <td>102.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>121.800000</td>\n      <td>0.000000</td>\n      <td>801.000000</td>\n      <td>594.000000</td>\n      <td>1.000000</td>\n      <td>2.330000</td>\n      <td>0.267000</td>\n    </tr>\n    <tr>\n      <th>25%</th>\n      <td>192.375000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>164.900000</td>\n      <td>0.000000</td>\n      <td>932.000000</td>\n      <td>730.950000</td>\n      <td>7.000000</td>\n      <td>23.710000</td>\n      <td>0.533000</td>\n    </tr>\n    <tr>\n      <th>50%</th>\n      <td>272.900000</td>\n      <td>22.000000</td>\n      <td>0.000000</td>\n      <td>185.000000</td>\n      <td>6.400000</td>\n      <td>968.000000</td>\n      <td>779.500000</td>\n      <td>28.000000</td>\n      <td>34.445000</td>\n      <td>0.675500</td>\n    </tr>\n    <tr>\n      <th>75%</th>\n      <td>350.000000</td>\n      <td>142.950000</td>\n      <td>118.300000</td>\n      <td>192.000000</td>\n      <td>10.200000</td>\n      <td>1029.400000</td>\n      <td>824.000000</td>\n      <td>56.000000</td>\n      <td>46.135000</td>\n      <td>0.935000</td>\n    </tr>\n    <tr>\n      <th>max</th>\n      <td>540.000000</td>\n      <td>359.400000</td>\n      <td>200.100000</td>\n      <td>247.000000</td>\n      <td>32.200000</td>\n      <td>1145.000000</td>\n      <td>992.600000</td>\n      <td>365.000000</td>\n      <td>82.600000</td>\n      <td>1.882000</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "            Cement  Blast Furnace Slag      Fly Ash        Water  \\\ncount  1030.000000         1030.000000  1030.000000  1030.000000   \nmean    281.167864           73.895825    54.188350   181.567282   \nstd     104.506364           86.279342    63.997004    21.354219   \nmin     102.000000            0.000000     0.000000   121.800000   \n25%     192.375000            0.000000     0.000000   164.900000   \n50%     272.900000           22.000000     0.000000   185.000000   \n75%     350.000000          142.950000   118.300000   192.000000   \nmax     540.000000          359.400000   200.100000   247.000000   \n\n       Superplasticizer  Coarse Aggregate  Fine Aggregate          Age  \\\ncount       1030.000000       1030.000000     1030.000000  1030.000000   \nmean           6.204660        972.918932      773.580485    45.662136   \nstd            5.973841         77.753954       80.175980    63.169912   \nmin            0.000000        801.000000      594.000000     1.000000   \n25%            0.000000        932.000000      730.950000     7.000000   \n50%            6.400000        968.000000      779.500000    28.000000   \n75%           10.200000       1029.400000      824.000000    56.000000   \nmax           32.200000       1145.000000      992.600000   365.000000   \n\n          Strength     WC Ratio  \ncount  1030.000000  1030.000000  \nmean     35.817961     0.748284  \nstd      16.705742     0.314000  \nmin       2.330000     0.267000  \n25%      23.710000     0.533000  \n50%      34.445000     0.675500  \n75%      46.135000     0.935000  \nmax      82.600000     1.882000  "
                    },
                    "execution_count": 7,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "# check out the descriptive statistics of the dataset \nconcrete_data.describe()"
        },
        {
            "cell_type": "code",
            "execution_count": 8,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": "Text(0, 0.5, 'Average')"
                    },
                    "execution_count": 8,
                    "metadata": {},
                    "output_type": "execute_result"
                },
                {
                    "data": {
                        "image/png": "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\n",
                        "text/plain": "<Figure size 1080x720 with 1 Axes>"
                    },
                    "metadata": {
                        "needs_background": "light"
                    },
                    "output_type": "display_data"
                }
            ],
            "source": "#create boxplot to visualize outliers\nplt.figure(figsize = (15,10))\nbox = sns.boxplot(data=concrete_data)\nplt.xlabel('Feature')\nplt.ylabel('Average')"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- The water, superplasricizer, fine aggregate, wc ratio, and age variables all contain some outliers so I decided to remove those outliers defined as the values that are 3 standard deviations (or more) away from the mean."
        },
        {
            "cell_type": "code",
            "execution_count": 9,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>Strength</th>\n      <th>WC Ratio</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>count</th>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.000000</td>\n      <td>971.00000</td>\n    </tr>\n    <tr>\n      <th>mean</th>\n      <td>279.192791</td>\n      <td>72.222142</td>\n      <td>57.070855</td>\n      <td>180.504531</td>\n      <td>6.246138</td>\n      <td>975.459320</td>\n      <td>775.267044</td>\n      <td>36.502575</td>\n      <td>35.466849</td>\n      <td>0.74243</td>\n    </tr>\n    <tr>\n      <th>std</th>\n      <td>101.721352</td>\n      <td>85.185068</td>\n      <td>64.483566</td>\n      <td>20.014525</td>\n      <td>5.485649</td>\n      <td>77.774221</td>\n      <td>76.980734</td>\n      <td>36.921353</td>\n      <td>16.688221</td>\n      <td>0.29621</td>\n    </tr>\n    <tr>\n      <th>min</th>\n      <td>116.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>121.800000</td>\n      <td>0.000000</td>\n      <td>801.000000</td>\n      <td>594.000000</td>\n      <td>1.000000</td>\n      <td>3.320000</td>\n      <td>0.28000</td>\n    </tr>\n    <tr>\n      <th>25%</th>\n      <td>192.000000</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>164.900000</td>\n      <td>0.000000</td>\n      <td>932.000000</td>\n      <td>739.000000</td>\n      <td>7.000000</td>\n      <td>23.455000</td>\n      <td>0.53350</td>\n    </tr>\n    <tr>\n      <th>50%</th>\n      <td>272.600000</td>\n      <td>20.000000</td>\n      <td>0.000000</td>\n      <td>183.800000</td>\n      <td>6.700000</td>\n      <td>970.000000</td>\n      <td>780.000000</td>\n      <td>28.000000</td>\n      <td>33.720000</td>\n      <td>0.67400</td>\n    </tr>\n    <tr>\n      <th>75%</th>\n      <td>349.500000</td>\n      <td>142.500000</td>\n      <td>118.300000</td>\n      <td>192.000000</td>\n      <td>10.300000</td>\n      <td>1040.000000</td>\n      <td>821.700000</td>\n      <td>28.000000</td>\n      <td>45.535000</td>\n      <td>0.93500</td>\n    </tr>\n    <tr>\n      <th>max</th>\n      <td>540.000000</td>\n      <td>316.100000</td>\n      <td>200.100000</td>\n      <td>228.000000</td>\n      <td>23.400000</td>\n      <td>1145.000000</td>\n      <td>992.600000</td>\n      <td>180.000000</td>\n      <td>82.600000</td>\n      <td>1.66000</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "           Cement  Blast Furnace Slag     Fly Ash       Water  \\\ncount  971.000000          971.000000  971.000000  971.000000   \nmean   279.192791           72.222142   57.070855  180.504531   \nstd    101.721352           85.185068   64.483566   20.014525   \nmin    116.000000            0.000000    0.000000  121.800000   \n25%    192.000000            0.000000    0.000000  164.900000   \n50%    272.600000           20.000000    0.000000  183.800000   \n75%    349.500000          142.500000  118.300000  192.000000   \nmax    540.000000          316.100000  200.100000  228.000000   \n\n       Superplasticizer  Coarse Aggregate  Fine Aggregate         Age  \\\ncount        971.000000        971.000000      971.000000  971.000000   \nmean           6.246138        975.459320      775.267044   36.502575   \nstd            5.485649         77.774221       76.980734   36.921353   \nmin            0.000000        801.000000      594.000000    1.000000   \n25%            0.000000        932.000000      739.000000    7.000000   \n50%            6.700000        970.000000      780.000000   28.000000   \n75%           10.300000       1040.000000      821.700000   28.000000   \nmax           23.400000       1145.000000      992.600000  180.000000   \n\n         Strength   WC Ratio  \ncount  971.000000  971.00000  \nmean    35.466849    0.74243  \nstd     16.688221    0.29621  \nmin      3.320000    0.28000  \n25%     23.455000    0.53350  \n50%     33.720000    0.67400  \n75%     45.535000    0.93500  \nmax     82.600000    1.66000  "
                    },
                    "execution_count": 9,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "#remove outliers defined as values greater than 3 STD from the mean\nz_scores = zscore(concrete_data) #z-score = (value-mean)/STD\nabs_z_scores = np.abs(z_scores)\nfiltered_entries = (abs_z_scores < 3).all(axis=1)\nconcrete_data = concrete_data[filtered_entries]\nconcrete_data.describe()"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Removing outliers got rid of 59 samples. However, most of the descriptive statistics for strength remain the same. Also, some rows with very high age were removed. This is okay because an older concrete will not be as representative of its original design strength."
        },
        {
            "cell_type": "code",
            "execution_count": 10,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": "Text(0.5, 1.0, 'Feature Correlation')"
                    },
                    "execution_count": 10,
                    "metadata": {},
                    "output_type": "execute_result"
                },
                {
                    "data": {
                        "image/png": "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\n",
                        "text/plain": "<Figure size 864x720 with 2 Axes>"
                    },
                    "metadata": {
                        "needs_background": "light"
                    },
                    "output_type": "display_data"
                }
            ],
            "source": "# check the correlation of the features \nplt.figure(figsize = (12, 10))\ncorr = concrete_data.corr()\nsns.heatmap(corr,annot=True).set_title(f'Feature Correlation')"
        },
        {
            "cell_type": "code",
            "execution_count": 11,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": "Cement                0.477993\nBlast Furnace Slag    0.161386\nFly Ash              -0.091692\nWater                -0.325331\nSuperplasticizer      0.377291\nCoarse Aggregate     -0.168868\nFine Aggregate       -0.149891\nAge                   0.432579\nStrength              1.000000\nWC Ratio             -0.496041\nName: Strength, dtype: float64"
                    },
                    "execution_count": 11,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "#display the correlation of each variable with strength \ncorr.Strength"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- No variables are that strongly correlated with strength so let's try to build a model with all of them.\n\n- Since WC Ratio was added, it could also be useful to remove the water and cement variables to improve model efficiency. But we can test this out later"
        },
        {
            "cell_type": "code",
            "execution_count": 12,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "image/png": "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\n",
                        "text/plain": "<Figure size 1080x1080 with 9 Axes>"
                    },
                    "metadata": {
                        "needs_background": "light"
                    },
                    "output_type": "display_data"
                }
            ],
            "source": "# visualize the distribution of the features \nplt.figure(figsize = (15, 15))\nfor idx,col in enumerate(concrete_data.columns,start=1):\n    if idx==8:\n        continue\n    elif idx>8:\n        idx=idx-1\n    plt.subplot(3,3,idx)\n    sns.histplot(data=concrete_data, x=col)"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Note for the variables of slag, fly ash, and superplasticizer, we have a lot of mixes with values of 0. In general, these components are the most expensive in concrete mixes and can be thought of as \"add-ons\", so it is reasonable that most mixes will not contain a lot of these components. I notice also that for fly ash there is not much data between 25-75 kg/m^3 and so I expect predictions for mixes wtih fly ash content in this range will not be very good."
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## A. Build a Baseline Regression Model with Keras "
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "Use the Keras library to build a neural network with the following:\n\n- One hidden layer with 10 nodes, and a ReLU activation function\n- Use the ADAM optimizer and the mean squared error as the loss function."
        },
        {
            "cell_type": "code",
            "execution_count": 13,
            "metadata": {},
            "outputs": [],
            "source": "# split the data into target and predictors \nconcrete_data_columns = concrete_data.columns\n# all columns except Strength\npredictors = concrete_data[concrete_data_columns[concrete_data_columns != 'Strength']] \ntarget = concrete_data['Strength'] # Strength column"
        },
        {
            "cell_type": "code",
            "execution_count": 14,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>WC Ratio</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1040.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>0.300</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>540.0</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>162.0</td>\n      <td>2.5</td>\n      <td>1055.0</td>\n      <td>676.0</td>\n      <td>28</td>\n      <td>0.300</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>266.0</td>\n      <td>114.0</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>670.0</td>\n      <td>90</td>\n      <td>0.857</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>380.0</td>\n      <td>95.0</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>594.0</td>\n      <td>28</td>\n      <td>0.600</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>266.0</td>\n      <td>114.0</td>\n      <td>0.0</td>\n      <td>228.0</td>\n      <td>0.0</td>\n      <td>932.0</td>\n      <td>670.0</td>\n      <td>28</td>\n      <td>0.857</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "   Cement  Blast Furnace Slag  Fly Ash  Water  Superplasticizer  \\\n0   540.0                 0.0      0.0  162.0               2.5   \n1   540.0                 0.0      0.0  162.0               2.5   \n5   266.0               114.0      0.0  228.0               0.0   \n7   380.0                95.0      0.0  228.0               0.0   \n8   266.0               114.0      0.0  228.0               0.0   \n\n   Coarse Aggregate  Fine Aggregate  Age  WC Ratio  \n0            1040.0           676.0   28     0.300  \n1            1055.0           676.0   28     0.300  \n5             932.0           670.0   90     0.857  \n7             932.0           594.0   28     0.600  \n8             932.0           670.0   28     0.857  "
                    },
                    "execution_count": 14,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "# check that the predictors are the ones we want \npredictors.head()"
        },
        {
            "cell_type": "code",
            "execution_count": 15,
            "metadata": {},
            "outputs": [
                {
                    "data": {
                        "text/plain": "0    79.99\n1    61.89\n5    47.03\n7    36.45\n8    45.85\nName: Strength, dtype: float64"
                    },
                    "execution_count": 15,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "# check that the target is the strength column\ntarget.head()"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Both predictor and strength variables look good!"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "For this baseline model, the data will not be normalized to check how the predictions are looking. For the last 2 models tested, the data will be normalized"
        },
        {
            "cell_type": "code",
            "execution_count": 16,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "Number of columns = 9\n"
                }
            ],
            "source": "# define the number of predictors\nn_cols = predictors.shape[1] \nprint(f'Number of columns = {n_cols}')"
        },
        {
            "cell_type": "code",
            "execution_count": 17,
            "metadata": {},
            "outputs": [],
            "source": "# let's define a function to create the regression neural network using keras \ndef regression_model():\n    # create model\n    model = Sequential()\n    #add one hidden layer with 10 nodes and ReLu activation function\n    model.add(Dense(10, activation='relu', input_shape=(n_cols,)))\n    model.add(Dense(1))\n    \n    # compile model\n    model.compile(optimizer='adam', loss='mean_squared_error')\n    return model"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "Let's define a function below to perform iterations of the model using a defined number of epochs and return the mean and standard deviation of the root mean squared error "
        },
        {
            "cell_type": "code",
            "execution_count": 18,
            "metadata": {},
            "outputs": [],
            "source": "def iterate_model(predictors,target,iterations,epochs):\n    clear_output(wait=True)\n    RMSE_by_iter = []\n    \n    for i in range(1,iterations+1):\n        # split the data into train set (70%) and test set (30%)\n        X_train, X_test, y_train, y_test = train_test_split(predictors, target, \\\n                                                         test_size=0.3, random_state=42)\n\n        # build the model\n        model = regression_model()\n        # fit the model and hold out 30% of the data for testing \n        history = model.fit(X_train, y_train, validation_data=(X_test, y_test), \\\n                             epochs=epochs, verbose=0)\n\n        # get predictions for strength \n        y_pred = model.predict(X_test)\n        #get the root mean squared error \n        RMSE = round(sqrt(MSE(y_test, y_pred)),3)\n        RMSE_by_iter.append(RMSE)\n        if i == 1 or i%5 == 0:\n            print(f'Iteration {i}: RMSE = {RMSE} MPa')\n            \n    mean_RMSE = round(mean(RMSE_by_iter),3)\n    std_RMSE = round(stdev(RMSE_by_iter),3)\n        \n    return mean_RMSE, std_RMSE"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- I chose to use root mean square error here since it is a little easier to quantify "
        },
        {
            "cell_type": "code",
            "execution_count": 19,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "Iteration 1: RMSE = 10.617 MPa\nIteration 5: RMSE = 23.034 MPa\nIteration 10: RMSE = 9.95 MPa\nIteration 15: RMSE = 9.754 MPa\nIteration 20: RMSE = 14.334 MPa\nIteration 25: RMSE = 12.506 MPa\nIteration 30: RMSE = 10.337 MPa\nIteration 35: RMSE = 39.712 MPa\nIteration 40: RMSE = 11.106 MPa\nIteration 45: RMSE = 9.765 MPa\nIteration 50: RMSE = 24.224 MPa\n\nAfter 50 iterations using 50 epochs, the mean_RMSE = 16.982 MPa and std_RMSE = 8.741 MPa\nThe model took 3.752 minutes to train\n"
                }
            ],
            "source": "# define the iterations and epochs and then train the network\niterations = 50\nepochs = 50 \n# train the model and time how long it takes \nstart = time.time()\nmean_RMSE, std_RMSE = iterate_model(predictors,target,iterations,epochs)\nend = time.time()\nprint(f'\\nAfter {iterations} iterations using {epochs} epochs, the mean_RMSE = {mean_RMSE} MPa and std_RMSE = {std_RMSE} MPa')\ntimer = round((end-start)/60,3)\nprint(f'The model took {timer} minutes to train')"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- The mean of the RMSE is around 17 MPa which is quite high and the standard deviation is 8.7 MPa. Let's see if normalizing the data, using more epochs, and more hidden layers will improve predictions."
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## B. Normalize the Data and Re-Train the Same Neural Network"
        },
        {
            "cell_type": "code",
            "execution_count": 20,
            "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>Cement</th>\n      <th>Blast Furnace Slag</th>\n      <th>Fly Ash</th>\n      <th>Water</th>\n      <th>Superplasticizer</th>\n      <th>Coarse Aggregate</th>\n      <th>Fine Aggregate</th>\n      <th>Age</th>\n      <th>WC Ratio</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>2.563938</td>\n      <td>-0.847826</td>\n      <td>-0.885045</td>\n      <td>-0.924555</td>\n      <td>-0.682898</td>\n      <td>0.829847</td>\n      <td>-1.289505</td>\n      <td>-0.230289</td>\n      <td>-1.493636</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2.563938</td>\n      <td>-0.847826</td>\n      <td>-0.885045</td>\n      <td>-0.924555</td>\n      <td>-0.682898</td>\n      <td>1.022713</td>\n      <td>-1.289505</td>\n      <td>-0.230289</td>\n      <td>-1.493636</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>-0.129695</td>\n      <td>0.490436</td>\n      <td>-0.885045</td>\n      <td>2.373050</td>\n      <td>-1.138632</td>\n      <td>-0.558788</td>\n      <td>-1.367447</td>\n      <td>1.448956</td>\n      <td>0.386784</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>0.991013</td>\n      <td>0.267393</td>\n      <td>-0.885045</td>\n      <td>2.373050</td>\n      <td>-1.138632</td>\n      <td>-0.558788</td>\n      <td>-2.354707</td>\n      <td>-0.230289</td>\n      <td>-0.480842</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>-0.129695</td>\n      <td>0.490436</td>\n      <td>-0.885045</td>\n      <td>2.373050</td>\n      <td>-1.138632</td>\n      <td>-0.558788</td>\n      <td>-1.367447</td>\n      <td>-0.230289</td>\n      <td>0.386784</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
                        "text/plain": "     Cement  Blast Furnace Slag   Fly Ash     Water  Superplasticizer  \\\n0  2.563938           -0.847826 -0.885045 -0.924555         -0.682898   \n1  2.563938           -0.847826 -0.885045 -0.924555         -0.682898   \n5 -0.129695            0.490436 -0.885045  2.373050         -1.138632   \n7  0.991013            0.267393 -0.885045  2.373050         -1.138632   \n8 -0.129695            0.490436 -0.885045  2.373050         -1.138632   \n\n   Coarse Aggregate  Fine Aggregate       Age  WC Ratio  \n0          0.829847       -1.289505 -0.230289 -1.493636  \n1          1.022713       -1.289505 -0.230289 -1.493636  \n5         -0.558788       -1.367447  1.448956  0.386784  \n7         -0.558788       -2.354707 -0.230289 -0.480842  \n8         -0.558788       -1.367447 -0.230289  0.386784  "
                    },
                    "execution_count": 20,
                    "metadata": {},
                    "output_type": "execute_result"
                }
            ],
            "source": "#normalize the predictor variables by subtracting the mean and dividing by standard deviation \npredictors_norm = (predictors - predictors.mean()) / predictors.std()\npredictors_norm.head()"
        },
        {
            "cell_type": "code",
            "execution_count": 21,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "Iteration 1: RMSE = 18.184 MPa\nIteration 5: RMSE = 18.337 MPa\nIteration 10: RMSE = 14.691 MPa\nIteration 15: RMSE = 20.036 MPa\nIteration 20: RMSE = 15.638 MPa\nIteration 25: RMSE = 18.503 MPa\nIteration 30: RMSE = 15.376 MPa\nIteration 35: RMSE = 15.317 MPa\nIteration 40: RMSE = 16.659 MPa\nIteration 45: RMSE = 22.139 MPa\nIteration 50: RMSE = 18.99 MPa\n\nAfter 50 iterations using 50 epochs, the mean_RMSE = 17.671 MPa and std_RMSE = 2.354 MPa\nThe model took 3.737 minutes to train\n"
                }
            ],
            "source": "#define the iterations and epochs and then train the network on the normalized predictors \niterations = 50\nepochs = 50 \n# train the model and time how long it takes \nstart = time.time()\nmean_RMSE, std_RMSE = iterate_model(predictors_norm,target,iterations,epochs)\nend = time.time()\nprint(f'\\nAfter {iterations} iterations using {epochs} epochs, the mean_RMSE = {mean_RMSE} MPa and std_RMSE = {std_RMSE} MPa')\ntimer = round((end-start)/60,3)\nprint(f'The model took {timer} minutes to train')"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Using the normalized predictors, the mean RMSE did not go down but the standard deviation did. Let's see what happens now if we increase the number of epochs to 100"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## C. Increse Epochs to 100 "
        },
        {
            "cell_type": "code",
            "execution_count": 22,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "Iteration 1: RMSE = 11.663 MPa\nIteration 5: RMSE = 11.822 MPa\nIteration 10: RMSE = 13.867 MPa\nIteration 15: RMSE = 12.236 MPa\nIteration 20: RMSE = 11.82 MPa\nIteration 25: RMSE = 11.784 MPa\nIteration 30: RMSE = 11.89 MPa\nIteration 35: RMSE = 12.104 MPa\nIteration 40: RMSE = 11.972 MPa\nIteration 45: RMSE = 12.029 MPa\nIteration 50: RMSE = 11.648 MPa\n\nAfter 50 iterations using 100 epochs, the mean_RMSE = 12.095 MPa and std_RMSE = 1.209 MPa\nThe model took 6.87 minutes to train\n"
                }
            ],
            "source": "#define the iterations and epochs and then train the network again\niterations = 50\nepochs = 100\n# train the model and time how long it takes \nstart = time.time()\nmean_RMSE, std_RMSE = iterate_model(predictors_norm,target,iterations,epochs)\nend = time.time()\nprint(f'\\nAfter {iterations} iterations using {epochs} epochs, the mean_RMSE = {mean_RMSE} MPa and std_RMSE = {std_RMSE} MPa')\ntimer = round((end-start)/60,3)\nprint(f'The model took {timer} minutes to train')"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- Using 100 epochs, the mean RMSE inproved by around 30% and the standard deviation dropped by around 50%. This is a significant improvement at a slight cost of extra compute time. Let's now see what happens if we add more hidden layers."
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "## D. Try 3 Hidden Layers and 50 Epochs "
        },
        {
            "cell_type": "code",
            "execution_count": 23,
            "metadata": {},
            "outputs": [],
            "source": "# let's re-define the regression model to include 3 hidden layers\ndef regression_model():\n    # create model\n    model = Sequential()\n    #add 3 hidden layers with 10 nodes and ReLu activation function\n    model.add(Dense(10, activation='relu', input_shape=(n_cols,)))\n    model.add(Dense(10, activation='relu', input_shape=(n_cols,)))\n    model.add(Dense(10, activation='relu', input_shape=(n_cols,)))\n    model.add(Dense(1))\n    \n    # compile model\n    model.compile(optimizer='adam', loss='mean_squared_error')\n    return model"
        },
        {
            "cell_type": "code",
            "execution_count": 24,
            "metadata": {},
            "outputs": [
                {
                    "name": "stdout",
                    "output_type": "stream",
                    "text": "Iteration 1: RMSE = 10.142 MPa\nIteration 5: RMSE = 11.073 MPa\nIteration 10: RMSE = 10.764 MPa\nIteration 15: RMSE = 10.383 MPa\nIteration 20: RMSE = 10.444 MPa\nIteration 25: RMSE = 10.916 MPa\nIteration 30: RMSE = 10.306 MPa\nIteration 35: RMSE = 10.485 MPa\nIteration 40: RMSE = 11.324 MPa\nIteration 45: RMSE = 10.516 MPa\nIteration 50: RMSE = 11.148 MPa\n\nAfter 50 iterations using 50 epochs, the mean_RMSE = 10.735 MPa and std_RMSE = 0.466 MPa\nThe model took 3.909 minutes to train\n"
                }
            ],
            "source": "#define the iterations and epochs and then train the network again\niterations = 50\nepochs = 50\n# train the model and time how long it takes \nstart = time.time()\nmean_RMSE, std_RMSE = iterate_model(predictors_norm,target,iterations,epochs)\nend = time.time()\nprint(f'\\nAfter {iterations} iterations using {epochs} epochs, the mean_RMSE = {mean_RMSE} MPa and std_RMSE = {std_RMSE} MPa')\ntimer = round((end-start)/60,3)\nprint(f'The model took {timer} minutes to train')"
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": "- This model is by far the best in terms of mean RMSE. In my opinion, however, this RMSE is still quite high and indicates that the average strength prediction is off by 10.7 MPa. \n\n- The standard deviation dropped quite significantly here meaning each iteration has roughly the same RMSE. This highlights the consistency of predictions in this nearual network with 3 hidden layers. \n\n- On top of this, the model did not take much longer to train than the previous models with just 1 hidden layer, which is good. "
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": ""
        }
    ],
    "metadata": {
        "kernelspec": {
            "display_name": "Python 3.8",
            "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": "3.8.11"
        }
    },
    "nbformat": 4,
    "nbformat_minor": 1
}