ML_Tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import needed packages\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this tutorial we will follow the Machine Learning Workflow to explore some data.  The data for this exercise is from the Iris data set, originally published at UCI Machine Learning Repository: Iris Data Set, this small dataset from 1936 is often used for testing out machine learning algorithms and visualizations. Each row of the table represents an iris flower, including its species and dimensions of its botanical parts, sepal and petal, in centimeters.\n",
    "\n",
    "<img src=\"img/ML_Workflow.png\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import scipy\n",
    "import numpy\n",
    "import matplotlib\n",
    "import pandas\n",
    "import sklearn\n",
    "from pandas.plotting import scatter_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import model_selection\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.svm import SVC"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load the data and explore the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load dataset\n",
    "url = \"https://raw.githubusercontent.com/maxgoff/ML_Example/master/iris.csv\"\n",
    "names = ['sepal-length', 'sepal-width', 'petal-length', 'petal-width', 'class']\n",
    "dataset = pandas.read_csv(url, names=names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(150, 5)\n"
     ]
    }
   ],
   "source": [
    "# shape\n",
    "print(dataset.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    sepal-length  sepal-width  petal-length  petal-width   class\n",
      "0            5.1          3.5           1.4          0.2  setosa\n",
      "1            4.9          3.0           1.4          0.2  setosa\n",
      "2            4.7          3.2           1.3          0.2  setosa\n",
      "3            4.6          3.1           1.5          0.2  setosa\n",
      "4            5.0          3.6           1.4          0.2  setosa\n",
      "5            5.4          3.9           1.7          0.4  setosa\n",
      "6            4.6          3.4           1.4          0.3  setosa\n",
      "7            5.0          3.4           1.5          0.2  setosa\n",
      "8            4.4          2.9           1.4          0.2  setosa\n",
      "9            4.9          3.1           1.5          0.1  setosa\n",
      "10           5.4          3.7           1.5          0.2  setosa\n",
      "11           4.8          3.4           1.6          0.2  setosa\n",
      "12           4.8          3.0           1.4          0.1  setosa\n",
      "13           4.3          3.0           1.1          0.1  setosa\n",
      "14           5.8          4.0           1.2          0.2  setosa\n",
      "15           5.7          4.4           1.5          0.4  setosa\n",
      "16           5.4          3.9           1.3          0.4  setosa\n",
      "17           5.1          3.5           1.4          0.3  setosa\n",
      "18           5.7          3.8           1.7          0.3  setosa\n",
      "19           5.1          3.8           1.5          0.3  setosa\n"
     ]
    }
   ],
   "source": [
    "# head\n",
    "print(dataset.head(20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       sepal-length  sepal-width  petal-length  petal-width\n",
      "count    150.000000   150.000000    150.000000   150.000000\n",
      "mean       5.843333     3.054000      3.758667     1.198667\n",
      "std        0.828066     0.433594      1.764420     0.763161\n",
      "min        4.300000     2.000000      1.000000     0.100000\n",
      "25%        5.100000     2.800000      1.600000     0.300000\n",
      "50%        5.800000     3.000000      4.350000     1.300000\n",
      "75%        6.400000     3.300000      5.100000     1.800000\n",
      "max        7.900000     4.400000      6.900000     2.500000\n"
     ]
    }
   ],
   "source": [
    "# descriptions\n",
    "print(dataset.describe())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visual the data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class\n",
      "setosa        50\n",
      "versicolor    50\n",
      "virginica     50\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# class distribution\n",
    "print(dataset.groupby('class').size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x142f6ee10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# box and whisker plots\n",
    "dataset.plot(kind='box', subplots=True, layout=(2,2), sharex=False, sharey=False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e4e3b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# histograms\n",
    "dataset.hist()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e4e3dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scatter plot matrix\n",
    "scatter_matrix(dataset)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate some algorithms\n",
    "Now we will create some models and estimate their accuracy.\n",
    "Covered in this step:\n",
    "    1. Separate out a validation dataset.\n",
    "    2. Set-up the test harness to use 10-fold cross validation.\n",
    "    3. Build 5 different models to predict species from flower measurements.\n",
    "    4. Select the best model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split-out validation dataset\n",
    "array = dataset.values\n",
    "X = array[:,0:4]\n",
    "Y = array[:,4]\n",
    "validation_size = 0.20\n",
    "seed = 7\n",
    "X_train, X_validation, Y_train, Y_validation = model_selection.train_test_split(X, Y, test_size=validation_size, random_state=seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Test options and evaluation metric\n",
    "seed = 7\n",
    "scoring = 'accuracy'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build models\n",
    "We don’t know which algorithms would be good on this problem or what configurations to use. We get an idea from the plots that some of the classes are partially linearly separable in some dimensions, so we are expecting generally good results.\n",
    "\n",
    "Let’s evaluate 6 different algorithms:\n",
    "\n",
    "- Logistic Regression (LR)\n",
    "- Linear Discriminant Analysis (LDA)\n",
    "- K-Nearest Neighbors (KNN).\n",
    "- Classification and Regression Trees (CART).\n",
    "- Gaussian Naive Bayes (NB).\n",
    "- Support Vector Machines (SVM)\n",
    "\n",
    "This is a good mixture of simple linear (LR and LDA), nonlinear (KNN, CART, NB and SVM) algorithms. We reset the random number seed before each run to ensure that the evaluation of each algorithm is performed using exactly the same data splits. It ensures the results are directly comparable.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Select the best model:\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/maxgoff/opt/miniconda3/envs/py3/lib/python3.6/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
      "/Users/maxgoff/opt/miniconda3/envs/py3/lib/python3.6/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n",
      "/Users/maxgoff/opt/miniconda3/envs/py3/lib/python3.6/site-packages/sklearn/linear_model/_logistic.py:765: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR: 0.983333 (0.033333)\n",
      "LDA: 0.975000 (0.038188)\n",
      "KNN: 0.983333 (0.033333)\n",
      "CART: 0.950000 (0.076376)\n",
      "NB: 0.966667 (0.040825)\n",
      "SVM: 0.983333 (0.033333)\n"
     ]
    }
   ],
   "source": [
    "# Spot Check Algorithms\n",
    "models = []\n",
    "models.append(('LR', LogisticRegression()))\n",
    "models.append(('LDA', LinearDiscriminantAnalysis()))\n",
    "models.append(('KNN', KNeighborsClassifier()))\n",
    "models.append(('CART', DecisionTreeClassifier()))\n",
    "models.append(('NB', GaussianNB()))\n",
    "models.append(('SVM', SVC()))\n",
    "# evaluate each model in turn\n",
    "results = []\n",
    "names = []\n",
    "print(\"Select the best model:\\n\")\n",
    "for name, model in models:\n",
    "\tkfold = model_selection.KFold(n_splits=10, shuffle=True, random_state=seed)\n",
    "\tcv_results = model_selection.cross_val_score(model, X_train, Y_train, cv=kfold, scoring=scoring)\n",
    "\tresults.append(cv_results)\n",
    "\tnames.append(name)\n",
    "\tmsg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std())\n",
    "\tprint(msg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1443f1ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compare Algorithms\n",
    "fig = plt.figure()\n",
    "fig.suptitle('Algorithm Comparison')\n",
    "ax = fig.add_subplot(111)\n",
    "plt.boxplot(results)\n",
    "ax.set_xticklabels(names)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Make Predictions\n",
    "The KNN algorithm was the most accurate model that we tested. Now we want to get an idea of the accuracy of the model on our validation set.\n",
    "\n",
    "This will give us an independent final check on the accuracy of the best model. It is valuable to keep a validation set just in case you made a slip during training, such as overfitting to the training set or a data leak. Both will result in an overly optimistic result.\n",
    "\n",
    "We can run the KNN model directly on the validation set and summarize the results as a final accuracy score, a confusion matrix and a classification report."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9\n",
      "[[ 7  0  0]\n",
      " [ 0 11  1]\n",
      " [ 0  2  9]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "      setosa       1.00      1.00      1.00         7\n",
      "  versicolor       0.85      0.92      0.88        12\n",
      "   virginica       0.90      0.82      0.86        11\n",
      "\n",
      "    accuracy                           0.90        30\n",
      "   macro avg       0.92      0.91      0.91        30\n",
      "weighted avg       0.90      0.90      0.90        30\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Make predictions on validation dataset\n",
    "knn = KNeighborsClassifier()\n",
    "knn.fit(X_train, Y_train)\n",
    "predictions = knn.predict(X_validation)\n",
    "print(accuracy_score(Y_validation, predictions))\n",
    "print(confusion_matrix(Y_validation, predictions))\n",
    "print(classification_report(Y_validation, predictions))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation\n",
    "\n",
    "How did we do?  Are the results good enough to move forward?  What changes would you make to improve the results?"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}