{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "colab": { "name": "MLAssignment.ipynb", "provenance": [], "collapsed_sections": [] }, "kernelspec": { "name": "python3", "display_name": "Python 3" }, "language_info": { "name": "python" } }, "cells": [ { "cell_type": "code", "metadata": { "id": "BZ-bXNuO_OyT" }, "source": [ "import pandas as pd\n", "import numpy as np\n", "from sklearn.model_selection import train_test_split\n", "from sklearn.naive_bayes import GaussianNB\n", "from sklearn import metrics" ], "execution_count": 36, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "fE4_SowlBrQV" }, "source": [ "dataset = pd.read_csv('sobar-72.csv')\n", "\n", "#Split to x and y\n", "\n", "y = dataset[\"ca_cervix\"]\n", "X = dataset.drop(\"ca_cervix\", 1)\n", "\n", "#filter the X to the different features\n", "x_behavior = dataset.filter(regex='behavior_')\n", "x_intention = dataset.filter(regex='intention_')\n", "x_attitude = dataset.filter(regex='attitude_')\n", "x_norm = dataset.filter(regex='norm_')\n", "x_perception = dataset.filter(regex='perception')\n", "x_motivation = dataset.filter(regex='motivation')\n", "x_socialSupport = dataset.filter(regex='socialSupport')\n", "x_empowerment = dataset.filter(regex='empowerment')\n" ], "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "cgjOou0ykm7A" }, "source": [ "#Behavior" ] }, { "cell_type": "code", "metadata": { "id": "8EzAEMpShjer", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "3bfd567d-d1cc-4aca-f815-a515565a3869" }, "source": [ "#Classifier for behavior features\n", "#Will be using Gaussian Naive Bayes classifier\n", "x_train, x_test, y_train, y_test = train_test_split(x_behavior, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 38, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.7333333333333333\n", " precision recall f1-score support\n", "\n", " 0 0.67 1.00 0.80 8\n", " 1 1.00 0.43 0.60 7\n", "\n", " accuracy 0.73 15\n", " macro avg 0.83 0.71 0.70 15\n", "weighted avg 0.82 0.73 0.71 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "5t8WTnsdkjlW" }, "source": [ "#Intention" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "bG1pvagXkf3u", "outputId": "6d133e1c-98ee-4981-b872-437d814cd474" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_intention, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 28, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.6666666666666666\n", " precision recall f1-score support\n", "\n", " 0 0.64 0.88 0.74 8\n", " 1 0.75 0.43 0.55 7\n", "\n", " accuracy 0.67 15\n", " macro avg 0.69 0.65 0.64 15\n", "weighted avg 0.69 0.67 0.65 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "ANoBrG_nlFy4" }, "source": [ "#Attitude" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "xIxqowb2lJDR", "outputId": "375ebcfe-1186-459d-984b-62d39c96a63c" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_attitude, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 39, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.5333333333333333\n", " precision recall f1-score support\n", "\n", " 0 0.53 1.00 0.70 8\n", " 1 0.00 0.00 0.00 7\n", "\n", " accuracy 0.53 15\n", " macro avg 0.27 0.50 0.35 15\n", "weighted avg 0.28 0.53 0.37 15\n", "\n" ], "name": "stdout" }, { "output_type": "stream", "text": [ "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1272: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n", " _warn_prf(average, modifier, msg_start, len(result))\n" ], "name": "stderr" } ] }, { "cell_type": "markdown", "metadata": { "id": "JU9k4xwq2dv7" }, "source": [ "#Norm" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "cuJGFNLdlgcd", "outputId": "c126cc3d-45b2-4cab-86b3-3efe2b29823a" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_norm, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 30, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.6666666666666666\n", " precision recall f1-score support\n", "\n", " 0 0.67 0.75 0.71 8\n", " 1 0.67 0.57 0.62 7\n", "\n", " accuracy 0.67 15\n", " macro avg 0.67 0.66 0.66 15\n", "weighted avg 0.67 0.67 0.66 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "z3P3_H8I2BYx" }, "source": [ "#Perception" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "NT_QTxYnmtvr", "outputId": "0721b601-4086-4993-db31-0e5beb3269ec" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_perception, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 31, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.8\n", " precision recall f1-score support\n", "\n", " 0 0.78 0.88 0.82 8\n", " 1 0.83 0.71 0.77 7\n", "\n", " accuracy 0.80 15\n", " macro avg 0.81 0.79 0.80 15\n", "weighted avg 0.80 0.80 0.80 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "bUaRGARt2D-m" }, "source": [ "#Motivation" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "R8NPlIlbnIAE", "outputId": "be426ad8-66a3-4869-e724-23f6ebbdcc22" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_motivation, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 32, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.7333333333333333\n", " precision recall f1-score support\n", "\n", " 0 0.67 1.00 0.80 8\n", " 1 1.00 0.43 0.60 7\n", "\n", " accuracy 0.73 15\n", " macro avg 0.83 0.71 0.70 15\n", "weighted avg 0.82 0.73 0.71 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "yskZntVt2GVA" }, "source": [ "#Social Support" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "jxwoOg3YnSd9", "outputId": "150ec548-0786-42cd-ea31-58bac1073671" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_socialSupport, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 33, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.6\n", " precision recall f1-score support\n", "\n", " 0 0.58 0.88 0.70 8\n", " 1 0.67 0.29 0.40 7\n", "\n", " accuracy 0.60 15\n", " macro avg 0.62 0.58 0.55 15\n", "weighted avg 0.62 0.60 0.56 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "i_c9zMME2Ip-" }, "source": [ "#Empowerment" ] }, { "cell_type": "code", "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "id": "qfGval8_na05", "outputId": "9d5d3cd6-1aea-404f-d7ee-efa1025791ca" }, "source": [ "x_train, x_test, y_train, y_test = train_test_split(x_empowerment, y, test_size=0.2, random_state=42)\n", "\n", "nb = GaussianNB()\n", "\n", "nb.fit(x_train, y_train)\n", "\n", "y_pred = nb.predict(x_test)\n", "\n", "print(\"Accuracy: \",metrics.accuracy_score(y_test, y_pred))\n", "print(metrics.classification_report(y_test, y_pred))" ], "execution_count": 34, "outputs": [ { "output_type": "stream", "text": [ "Accuracy: 0.7333333333333333\n", " precision recall f1-score support\n", "\n", " 0 0.70 0.88 0.78 8\n", " 1 0.80 0.57 0.67 7\n", "\n", " accuracy 0.73 15\n", " macro avg 0.75 0.72 0.72 15\n", "weighted avg 0.75 0.73 0.73 15\n", "\n" ], "name": "stdout" } ] }, { "cell_type": "markdown", "metadata": { "id": "JIIrUmET1JZs" }, "source": [ "All the classifirs above are GaussianNaive Bayes classifiers. Each classifier runs on data from a different feature from the behaviour risk cervical cancer dataset. These unique features are: behavior, intention, attitude, norm, perception, motivation, social suppport and empowerment. From the above runs, the feature with the best F1 accuracy score is **Perception** with an accuracy score of 0.8. Meaning the classifier predicted correctly the presence of cervical cancer in patients 80% of the time.\n", "Behaviour, motivation and empowerment features all had an accuracy of 73%, meaning they are features with similar characteristics and can be utilised by the classifier interchangeably, same case for norm and intention with accuracy scores of 66%. The worst performers are Social Support with 60% and attitude with 53%. \n", "This also shows that there are 4 distinct features in the data that can be utilized by classification algorithms." ] } ] }