{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Interpreting Anomalies from Isolation Forest\n", "\n", "\n", "## Isolation Forest\n", "\n", "The idea behind Isolation Forest is that anomalies are easier to separate from the rest of the data than other points. The Isolation Forest algorithm partitions the data through a forest of decision trees. Each split is made randomly. The number of splits it takes to isolate a record indicates whether or not the record is an anomaly. When a forest of random trees collectively produces shorter path lengths for particular samples, they are highly likely to be anomalies.\n", "\n", "In this demo, we will use the Isolation Forest technique to find employees that may be anomalies. \n", "\n", "\n", "## Loading the Data\n", "\n", "Before we dive into the anomaly detection, let's initialize the h2o cluster and load our data in. We will be using the [synthetic employee attrition dataset](https://www.ibm.com/communities/analytics/watson-analytics-blog/hr-employee-attrition/). This contains a record per employee with information about their employment history and whether they engaged in attrition." ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n", "Attempting to start a local H2O server...\n", " Java Version: java version \"1.8.0_91\"; Java(TM) SE Runtime Environment (build 1.8.0_91-b14); Java HotSpot(TM) 64-Bit Server VM (build 25.91-b14, mixed mode)\n", " Starting server from /Users/megankurka/anaconda3/lib/python3.6/site-packages/h2o/backend/bin/h2o.jar\n", " Ice root: /var/folders/gv/w2f3zs_d33l3dt5j67k9nhhr0000gn/T/tmpl241x2j4\n", " JVM stdout: /var/folders/gv/w2f3zs_d33l3dt5j67k9nhhr0000gn/T/tmpl241x2j4/h2o_megankurka_started_from_python.out\n", " JVM stderr: /var/folders/gv/w2f3zs_d33l3dt5j67k9nhhr0000gn/T/tmpl241x2j4/h2o_megankurka_started_from_python.err\n", " Server is running at http://127.0.0.1:54321\n", "Connecting to H2O server at http://127.0.0.1:54321... successful.\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
H2O cluster uptime:01 secs
H2O cluster timezone:America/Los_Angeles
H2O data parsing timezone:UTC
H2O cluster version:3.22.1.2
H2O cluster version age:4 days
H2O cluster name:H2O_from_python_megankurka_77x8qj
H2O cluster total nodes:1
H2O cluster free memory:3.556 Gb
H2O cluster total cores:8
H2O cluster allowed cores:8
H2O cluster status:accepting new members, healthy
H2O connection url:http://127.0.0.1:54321
H2O connection proxy:None
H2O internal security:False
H2O API Extensions:XGBoost, Algos, AutoML, Core V3, Core V4
Python version:3.6.0 final
" ], "text/plain": [ "-------------------------- ----------------------------------------\n", "H2O cluster uptime: 01 secs\n", "H2O cluster timezone: America/Los_Angeles\n", "H2O data parsing timezone: UTC\n", "H2O cluster version: 3.22.1.2\n", "H2O cluster version age: 4 days\n", "H2O cluster name: H2O_from_python_megankurka_77x8qj\n", "H2O cluster total nodes: 1\n", "H2O cluster free memory: 3.556 Gb\n", "H2O cluster total cores: 8\n", "H2O cluster allowed cores: 8\n", "H2O cluster status: accepting new members, healthy\n", "H2O connection url: http://127.0.0.1:54321\n", "H2O connection proxy:\n", "H2O internal security: False\n", "H2O API Extensions: XGBoost, Algos, AutoML, Core V3, Core V4\n", "Python version: 3.6.0 final\n", "-------------------------- ----------------------------------------" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import h2o\n", "h2o.init()" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Parse progress: |█████████████████████████████████████████████████████████| 100%\n" ] } ], "source": [ "employee_data = h2o.import_file(\"./WA_Fn-UseC_-HR-Employee-Attrition.csv\")" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
AgeAttrition BusinessTravel DailyRateDepartment DistanceFromHome EducationEducationField EmployeeCount EmployeeNumber EnvironmentSatisfactionGender HourlyRate JobInvolvement JobLevelJobRole JobSatisfactionMaritalStatus MonthlyIncome MonthlyRate NumCompaniesWorkedOver18 OverTime PercentSalaryHike PerformanceRating RelationshipSatisfaction StandardHours StockOptionLevel TotalWorkingYears TrainingTimesLastYear WorkLifeBalance YearsAtCompany YearsInCurrentRole YearsSinceLastPromotion YearsWithCurrManager
41Yes Travel_Rarely 1102Sales 1 2Life Sciences 1 1 2Female 94 3 2Sales Executive 4Single 5993 19479 8Y Yes 11 3 1 80 0 8 0 1 6 4 0 5
49No Travel_Frequently 279Research & Development 8 1Life Sciences 1 2 3Male 61 2 2Research Scientist 2Married 5130 24907 1Y No 23 4 4 80 1 10 3 3 10 7 1 7
37Yes Travel_Rarely 1373Research & Development 2 2Other 1 4 4Male 92 2 1Laboratory Technician 3Single 2090 2396 6Y Yes 15 3 2 80 0 7 3 3 0 0 0 0
33No Travel_Frequently 1392Research & Development 3 4Life Sciences 1 5 4Female 56 3 1Research Scientist 3Married 2909 23159 1Y Yes 11 3 3 80 0 8 3 3 8 7 3 0
27No Travel_Rarely 591Research & Development 2 1Medical 1 7 1Male 40 3 1Laboratory Technician 2Married 3468 16632 9Y No 12 3 4 80 1 6 3 3 2 2 2 2
32No Travel_Frequently 1005Research & Development 2 2Life Sciences 1 8 4Male 79 3 1Laboratory Technician 4Single 3068 11864 0Y No 13 3 3 80 0 8 2 2 7 7 3 6
59No Travel_Rarely 1324Research & Development 3 3Medical 1 10 3Female 81 4 1Laboratory Technician 1Married 2670 9964 4Y Yes 20 4 1 80 3 12 3 2 1 0 0 0
30No Travel_Rarely 1358Research & Development 24 1Life Sciences 1 11 4Male 67 3 1Laboratory Technician 3Divorced 2693 13335 1Y No 22 4 2 80 1 1 2 3 1 0 0 0
38No Travel_Frequently 216Research & Development 23 3Life Sciences 1 12 4Male 44 2 3Manufacturing Director 3Single 9526 8787 0Y No 21 4 2 80 0 10 2 3 9 7 1 8
36No Travel_Rarely 1299Research & Development 27 3Medical 1 13 3Male 94 3 2Healthcare Representative 3Married 5237 16577 6Y No 13 3 2 80 2 17 3 2 7 7 7 7
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "employee_data.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Isolation Forests\n", "\n", "To find our anomalous employees, let's train our isolation forest and see how the predictions look. We will only use a subset of columns for demo purposes." ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "isolationforest Model Build progress: |███████████████████████████████████| 100%\n" ] } ], "source": [ "from h2o.estimators import H2OIsolationForestEstimator\n", "myX = ['Age', 'BusinessTravel', 'DistanceFromHome', 'Education', 'Gender', 'JobInvolvement', 'JobLevel', \n", " 'MaritalStatus', 'MonthlyIncome', 'NumCompaniesWorked', 'OverTime', 'PercentSalaryHike',\n", " 'PerformanceRating', 'TotalWorkingYears', 'TrainingTimesLastYear', 'YearsAtCompany', \n", " 'YearsInCurrentRole', 'YearsSinceLastPromotion', 'YearsWithCurrManager']\n", "\n", "isolation_model = H2OIsolationForestEstimator(model_id = \"isolation_forest.hex\", seed = 1234)\n", "isolation_model.train(training_frame = employee_data, x = myX)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The predictions from the isolation forest return the `mean_length`. This is the average number of splits it took to isolate the record across all the decision trees in the forest. Records with a smaller `mean_length` are more likely to be anomalous since it takes fewer partitions of the data to isolate them." ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "isolationforest prediction progress: |████████████████████████████████████| 100%\n" ] }, { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
predict mean_length
0.120301 6.54
0.037594 6.76
0.180451 6.38
-0.0225564 6.92
-0.0150376 6.9
-0.0451128 6.98
0.172932 6.4
0.0225564 6.8
0.150376 6.46
0.037594 6.76
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "predictions = isolation_model.predict(employee_data)\n", "predictions.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The histogram of the `mean_length` shows that most employees have a `mean_length` greater than 6.5. This means that it takes more than 6 splits on average to partition them. " ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY8AAAEXCAYAAABVr8jJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n5glzSVJnhockqTPDQ5LUmeEhSers/wPOXYYUHn/M1QAAAABJRU5ErkJggg==\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "predictions[\"mean_length\"].hist()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Defining Anomalies\n", "\n", "We will define an anomaly as an employee who's `mean_length` is less than 5.5. These were employees who were easier to isolate from the rest of the data.\n", "\n", "There are 34 anomalous employees." ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Number of Anomalies: 34\n" ] } ], "source": [ "anomalies = employee_data[predictions[\"mean_length\"] < 5.5]\n", "print(\"Number of Anomalies: \" + str(anomalies.nrow))" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "isolationforest prediction progress: |████████████████████████████████████| 100%\n" ] }, { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
mean_length AgeBusinessTravel DistanceFromHome EducationGender JobInvolvement JobLevelMaritalStatus MonthlyIncome NumCompaniesWorkedOverTime PercentSalaryHike PerformanceRating TotalWorkingYears TrainingTimesLastYear YearsAtCompany YearsInCurrentRole YearsSinceLastPromotion YearsWithCurrManager
5.48 59Travel_Rarely 25 3Female 3 3Single 7637 7No 11 3 28 3 21 16 7 9
5.38 58Travel_Rarely 10 4Male 3 4Single 13872 0No 13 3 38 1 37 10 1 8
5.4 59Non-Travel 2 4Female 2 5Married 18844 9No 21 4 30 3 3 2 2 2
4.88 51Travel_Rarely 6 3Male 3 5Single 19537 7No 13 3 23 5 20 18 15 15
4.92 58Travel_Rarely 23 4Female 3 3Married 10312 1No 12 3 40 3 40 10 15 6
5.48 46Travel_Rarely 1 2Female 3 3Divorced 10453 1No 25 4 24 2 24 13 15 7
4.92 55Travel_Rarely 14 4Male 4 5Single 18722 8No 11 3 36 3 24 15 2 15
5.04 52Travel_Rarely 1 4Male 2 5Married 19999 0No 14 3 34 5 33 18 11 9
5.04 52Non-Travel 2 4Male 2 5Single 19068 1Yes 18 3 33 2 33 7 15 12
4.92 55Travel_Rarely 1 3Male 3 5Single 19045 0Yes 14 3 37 2 36 10 4 13
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "isolation_model.predict(anomalies)[\"mean_length\"].cbind(anomalies[myX])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Interpreting Anomalies\n", "\n", "There are two levels of interpretation:\n", "\n", "* global level: high level understanding of what segments of data are considered anomalous\n", "* local level: understanding of why an individual record is considered anomalous\n", "\n", "We will start with the global level. Our goal is to gain an understanding of what segments of data are considered anomalous.\n", "\n", "### Global Level\n", "\n", "Now that we have found anomalous employees, we are interested in why they are considered anomalies. To do this, we will train a surrogate decision tree. The purpose of the surrogate decision tree is to find records with the anomaly flag. To do this, it will find segments of similar anomalies and discover how to separate them from records that are not anomalies. We can use this decision tree to then describe anomalous segments of the data.\n", "\n", "The steps of interpreting anomalies on a global level are:\n", "\n", "1. Create a frame with a column that indicates whether the record was considered an anomaly.\n", "2. Train a decision tree to predict the anomaly flag.\n", "3. Visualize the decision tree to determine which segments of the data are considered anomalous.\n", "\n", "In our first step, we will add a column called `anomaly`. This is a flag that indicates whether the isolation forest considered the record an anomaly." ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
anomaly Count
No 1436
Yes 34
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "global_surrogate_data = employee_data[:, :]\n", "global_surrogate_data[\"anomaly\"] = (predictions[\"mean_length\"] < 5.5).ifelse(\"Yes\", \"No\")\n", "global_surrogate_data[\"anomaly\"].table()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now that we have the surrogate data, we can train a single decision tree to predict the anomaly flag. We will keep this decision tree simple (only a single decision tree with a depth of 3) because the purpose of the decision tree is to be completely interpretable." ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "drf Model Build progress: |███████████████████████████████████████████████| 100%\n" ] } ], "source": [ "from h2o.estimators import H2ORandomForestEstimator\n", "\n", "global_surrogate_dt = H2ORandomForestEstimator(model_id = \"global_surrogate_decision_tree.hex\", \n", " ntrees = 1, max_depth = 3,\n", " sample_rate = 1, mtries = len(myX))\n", "global_surrogate_dt.train(training_frame = global_surrogate_data, x = myX, y = \"anomaly\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can now visualize the decision tree to find segments of the data that are anomalous." ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "collapsed": true }, "outputs": [], "source": [ "import os\n", "import subprocess\n", "from IPython.display import Image\n", "def generateTreeImage(decision_tree, image_file_path):\n", " # Download MOJO\n", " mojo_path = decision_tree.download_mojo(get_genmodel_jar=True)\n", " directory = os.path.dirname(mojo_path)\n", " h2o_jar_path = os.path.join(directory, \"h2o-genmodel.jar\")\n", " # Create Graphviz file\n", " gv_file_path = os.path.join(directory, \"decision_tree.gv\")\n", " gv_call = \" \".join([\"java\", \"-cp\", h2o_jar_path, \"hex.genmodel.tools.PrintMojo\", \"--tree 0 -i\", mojo_path , \"-o\", gv_file_path])\n", " result = subprocess.call(gv_call, shell=True)\n", " result = subprocess.call([\"ls\", gv_file_path], shell = False)\n", " result = subprocess.call([\"dot\", \"-Tpng\", gv_file_path, \"-o\", image_file_path], shell=False)\n", " result = subprocess.call([\"ls\",image_file_path], shell = False)\n", " \n", " return Image(image_file_path)" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "generateTreeImage(global_surrogate_dt, \"./global_surrogate_decision_tree.png\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The visualization shows our global surrogate decision tree. The values in the leaf nodes represent the probability of the record not being an anomaly. We are, therefore, interested in leaf nodes with low values - these will indicate a segment of data that is anomalous.\n", "\n", "We can see that there are three leaf nodes with all anomalies. One leaf node is defined as: \n", "\n", "* Total Working Years < 32.5\n", "* Years Since Last Promotion < 12.5\n", "* Percent Salary Hike >= 18.0\n", "\n", "This segment seems strange for two reasons: \n", "\n", "* the employees have been working 32 years or less but have not received a promotion for more than 12 years\n", " * most employees have a promotion ever 4 years of working \n", "* the employee has not had a promotion recently but has received a large salary hike\n", " * years since last promotion is negatively correlated with salary hike" ] }, { "cell_type": "code", "execution_count": 47, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/plain": [ "[4.0]" ] }, "execution_count": 47, "metadata": {}, "output_type": "execute_result" } ], "source": [ "promotion_per_working_years = employee_data[employee_data[\"YearsSinceLastPromotion\"] > 0]\n", "promotion_per_working_years = promotion_per_working_years[\"TotalWorkingYears\"]/promotion_per_working_years[\"YearsSinceLastPromotion\"]\n", "promotion_per_working_years.median()" ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
YearsSinceLastPromotion PercentSalaryHike
1 -0.0221543
-0.0221543 1
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "employee_data[[\"YearsSinceLastPromotion\", \"PercentSalaryHike\"]].cor()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Local Level\n", "\n", "Now we will perform a local level interpretation. The goal of this interpretation is to determine why a specific employee is considered an anomaly.\n", "\n", "The steps of interpreting anomalies on a local level are:\n", "\n", "1. Create a frame with a column that indicates whether the record is our selected anomaly.\n", "2. Train a decision tree to predict the anomaly flag.\n", "3. Visualize the decision tree to determine how the selected anomaly separates from the rest of the dat.a.\n", "\n", "Let's begin by examining our first anomaly." ] }, { "cell_type": "code", "execution_count": 48, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
AgeBusinessTravel DistanceFromHome EducationGender JobInvolvement JobLevelMaritalStatus MonthlyIncome NumCompaniesWorkedOverTime PercentSalaryHike PerformanceRating TotalWorkingYears TrainingTimesLastYear YearsAtCompany YearsInCurrentRole YearsSinceLastPromotion YearsWithCurrManager
59Travel_Rarely 25 3Female 3 3Single 7637 7No 11 3 28 3 21 16 7 9
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 48, "metadata": {}, "output_type": "execute_result" } ], "source": [ "anomalies[0, myX]" ] }, { "cell_type": "code", "execution_count": 49, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "isolationforest prediction progress: |████████████████████████████████████| 100%\n" ] }, { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
predict mean_length
0.518797 5.48
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 49, "metadata": {}, "output_type": "execute_result" } ], "source": [ "isolation_model.predict(anomalies[0, :])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To determine why this employee is considered anomalous, we will build a surrogate decision tree. The goal of the decision tree is to separate this employee from all other employees.\n", "\n", "The structure of the decision tree will tell us why the employee is different from others." ] }, { "cell_type": "code", "execution_count": 54, "metadata": { "collapsed": true }, "outputs": [], "source": [ "local_surrogate_data = employee_data[:, :]\n", "local_surrogate_data[\"anomaly_record\"] = (local_surrogate_data[\"EmployeeNumber\"] == 81).ifelse(\"Anomaly\", \"NotAnomaly\")" ] }, { "cell_type": "code", "execution_count": 55, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
anomaly_record Count
NotAnomaly 1469
Anomaly 1
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 55, "metadata": {}, "output_type": "execute_result" } ], "source": [ "local_surrogate_data[\"anomaly_record\"].table()" ] }, { "cell_type": "code", "execution_count": 56, "metadata": { "collapsed": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "drf Model Build progress: |███████████████████████████████████████████████| 100%\n" ] } ], "source": [ "local_surrogate_dt = H2ORandomForestEstimator(model_id = \"local_level_surrogate_decision_tree.hex\", \n", " ntrees = 1, max_depth = 3,\n", " sample_rate = 1, mtries = len(myX))\n", "local_surrogate_dt.train(training_frame = local_surrogate_data, x = myX, y = \"anomaly_record\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can visualize this decision tree to see how it split to isolate our anomaly record." ] }, { "cell_type": "code", "execution_count": 57, "metadata": { "collapsed": false, "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "" ] }, "execution_count": 57, "metadata": {}, "output_type": "execute_result" } ], "source": [ "generateTreeImage(local_surrogate_dt, \"./global_surrogate_decision_tree.png\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The anomalous employee falls in the bucket of employees with a high number of years in the Current Role and Age. It falls in the bucket: `YearsInCurrentRole >= 15.5` and `Age >= 57.5`. \n", "\n", "We can see that our simple decision tree is perfectly able to separate the anomaly from the other employees because it has an AUC of 1. This means that this employee is the only one in the data that has been in his/her current role more than 15.5 years and is older than 57." ] }, { "cell_type": "code", "execution_count": 58, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/plain": [ "1.0" ] }, "execution_count": 58, "metadata": {}, "output_type": "execute_result" } ], "source": [ "local_surrogate_dt.model_performance(surrogate_data).auc()" ] }, { "cell_type": "code", "execution_count": 59, "metadata": { "collapsed": false }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n", "\n", "\n", "\n", "
Age YearsInCurrentRole
59 16
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [] }, "execution_count": 59, "metadata": {}, "output_type": "execute_result" } ], "source": [ "anomalies[0, [\"Age\", \"YearsInCurrentRole\"]]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "If we examine the distribution of these two features, we can see that the employee falls on the right of the spectrum for both." ] }, { "cell_type": "code", "execution_count": 60, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkCS1MiAkSa3+P9l5otXi7A5yAAAAAElFTkSuQmCC\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "employee_data[\"Age\"].hist()" ] }, { "cell_type": "code", "execution_count": 61, "metadata": { "collapsed": false }, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "employee_data[\"YearsInCurrentRole\"].hist()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": true }, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "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": "3.6.0" } }, "nbformat": 4, "nbformat_minor": 2 }