Upload Fraud_Detection_AutoML.ipynb
Browse files
Fraud_Detection_AutoML.ipynb
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Fraud Detection AutoML.ipynb","provenance":[],"authorship_tag":"ABX9TyNEt/sTUkfLrRq3rL38Ef0z"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"code","execution_count":1,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"zCL6MsfRcvaY","executionInfo":{"status":"ok","timestamp":1649725775957,"user_tz":240,"elapsed":5052,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}},"outputId":"8ac2429f-e4e1-46ff-85e6-37e8d210579a"},"outputs":[{"output_type":"stream","name":"stdout","text":["Requirement already satisfied: scipy==1.7.0 in /usr/local/lib/python3.7/dist-packages (1.7.0)\n","Requirement already satisfied: numpy<1.23.0,>=1.16.5 in /usr/local/lib/python3.7/dist-packages (from scipy==1.7.0) (1.21.5)\n"]}],"source":["!pip install auto-sklearn\n","!pip install scipy==1.7.0"]},{"cell_type":"code","source":["# data and metric imports\n","import sklearn.model_selection\n","import sklearn.metrics\n","\n","import pandas as pd\n","df = pd.read_csv('Dataset.csv')\n","print(df)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"510DiiSndEh5","executionInfo":{"status":"ok","timestamp":1649726251148,"user_tz":240,"elapsed":1306,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}},"outputId":"16809038-2b8b-4537-8e73-935fb47f61f5"},"execution_count":10,"outputs":[{"output_type":"stream","name":"stdout","text":[" BELNR WAERS BUKRS KTOSL PRCTR BSCHL HKONT DMBTR WRBTR \\\n","0 288203 C3 C31 C9 C92 A3 B1 280979.60 0.00 \n","1 324441 C1 C18 C7 C76 A1 B2 129856.53 243343.00 \n","2 133537 C1 C19 C2 C20 A1 B3 957463.97 3183838.41 \n","3 331521 C4 C48 C9 C95 A2 B1 2681709.51 28778.00 \n","4 375333 C5 C58 C1 C19 A3 B1 910514.49 346.00 \n","... ... ... ... ... ... ... ... ... ... \n","533004 446818 C1 C18 C3 C32 A1 B2 2501589.15 0.00 \n","533005 455564 C8 C80 C1 C11 A1 B1 390076.18 12065.45 \n","533006 156896 C1 C10 C1 C19 A1 B1 192147.38 326823.09 \n","533007 455245 C1 C14 C4 C40 A1 B3 870539.68 0.00 \n","533008 281395 C8 C80 C7 C78 A1 B2 532106.87 32467.08 \n","\n"," label \n","0 regular \n","1 regular \n","2 regular \n","3 regular \n","4 regular \n","... ... \n","533004 regular \n","533005 regular \n","533006 regular \n","533007 regular \n","533008 regular \n","\n","[533009 rows x 10 columns]\n"]}]},{"cell_type":"code","source":["df.iloc[:, 1:7] = df.iloc[:, 1:7].astype('category')\n","df.iloc[:, 9] = df.iloc[:, 9].astype('category')"],"metadata":{"id":"9luBWV2xhW0A","executionInfo":{"status":"ok","timestamp":1649726251868,"user_tz":240,"elapsed":724,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}}},"execution_count":11,"outputs":[]},{"cell_type":"code","source":["df.dtypes"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"V_ArN_Eghlxh","executionInfo":{"status":"ok","timestamp":1649726251869,"user_tz":240,"elapsed":6,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}},"outputId":"7e2d9163-a4bc-4008-ba3f-a4aed05e675b"},"execution_count":12,"outputs":[{"output_type":"execute_result","data":{"text/plain":["BELNR int64\n","WAERS category\n","BUKRS category\n","KTOSL category\n","PRCTR category\n","BSCHL category\n","HKONT category\n","DMBTR float64\n","WRBTR float64\n","label category\n","dtype: object"]},"metadata":{},"execution_count":12}]},{"cell_type":"code","source":["X = df.iloc[:, :9]\n","y = df.iloc[:, 9]\n","X_train, X_test, y_train, y_test = \\\n"," sklearn.model_selection.train_test_split(X, y, random_state=1)"],"metadata":{"id":"IHuCaInWdrkl","executionInfo":{"status":"ok","timestamp":1649726252408,"user_tz":240,"elapsed":542,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}}},"execution_count":13,"outputs":[]},{"cell_type":"code","source":["import autosklearn.classification\n","cls = autosklearn.classification.AutoSklearnClassifier()\n","cls.fit(X_train, y_train)\n","predictions = cls.predict(X_test)\n","print(\"Accuracy Score\", sklearn.metrics.accuracy_score(y_test, predictions))"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"qNn1CXKxdFFx","executionInfo":{"status":"ok","timestamp":1649729927505,"user_tz":240,"elapsed":3675103,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}},"outputId":"96072b18-395d-4ba1-9e8e-23876e603890"},"execution_count":14,"outputs":[{"output_type":"stream","name":"stdout","text":["[WARNING] [2022-04-12 02:03:40,372:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:04:09,099:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:04:33,902:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:05:05,464:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:05:35,901:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:05:47,006:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:11:48,685:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:13:33,556:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:13:44,399:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:14:10,272:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:14:20,249:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","[WARNING] [2022-04-12 02:14:47,846:smac.runhistory.runhistory2epm.RunHistory2EPM4LogCost] Got cost of smaller/equal to 0. Replace by 0.000010 since we use log cost.\n","Accuracy Score 1.0\n"]}]},{"cell_type":"code","source":["from sklearn.metrics import confusion_matrix\n","confusion_matrix(y_test, predictions)\n","\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","from sklearn import metrics\n","cm = metrics.confusion_matrix(y_test, predictions)\n","plt.figure(figsize=(9,9))\n","sns.heatmap(cm, annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'Blues_r');\n","plt.ylabel('Actual label');\n","plt.xlabel('Predicted label');\n","all_sample_title = 'Accuracy Score: {0}'.format(sklearn.metrics.accuracy_score(y_test, predictions))\n","plt.title(all_sample_title, size = 15);"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":526},"id":"PRtV6453ViKB","executionInfo":{"status":"ok","timestamp":1649729929085,"user_tz":240,"elapsed":1590,"user":{"displayName":"Kyle Masiak","userId":"03455281970866158381"}},"outputId":"078ce159-057e-4cd3-e22f-d1b4d8858874"},"execution_count":15,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 648x648 with 2 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]}]}
|