{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"mount_file_id":"1Fw3GjFHrS_S_0umDDOEqO0kTRJ9tavQX","authorship_tag":"ABX9TyPdJ0Elmyst3XcwdnXfsiOn"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"markdown","source":["# **Studi Kasus Heart Disease 5 Fitur Part 2**"],"metadata":{"id":"s8B30LBor6qp"}},{"cell_type":"markdown","source":["Implementasi dengan Menggunakan Model `Bagging Classifier` , `Random forest` dan `Stacking clasifier`"],"metadata":{"id":"9CEPTpErkvJw"}},{"cell_type":"markdown","source":["## Membaca data"],"metadata":{"id":"EAF-vi3ksEd5"}},{"cell_type":"code","execution_count":38,"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":206},"id":"9xrzLAOwrx-Q","executionInfo":{"status":"ok","timestamp":1669557916548,"user_tz":-420,"elapsed":423,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"95298aa9-3240-40cf-832e-382ad2e12e97"},"outputs":[{"output_type":"execute_result","data":{"text/plain":[" age sex BP cholestrol heart disease\n","0 70 1 130 322 1\n","1 67 0 115 564 0\n","2 57 1 124 261 1\n","3 64 1 128 263 0\n","4 74 0 120 269 0"],"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"," \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","
agesexBPcholestrolheart disease
07011303221
16701155640
25711242611
36411282630
47401202690
\n","
\n"," \n"," \n"," \n","\n"," \n","
\n","
\n"," "]},"metadata":{},"execution_count":38}],"source":["from scipy.io import arff\n","import pandas as pd\n","from sklearn.preprocessing import MinMaxScaler\n","import joblib\n","\n","data = pd.read_csv('https://raw.githubusercontent.com/soumya-mishra/Heart-Disease_DT/main/heart_v2.csv')\n","df = data\n","df.head()"]},{"cell_type":"markdown","source":["### Memisahkan Label"],"metadata":{"id":"o1L0ks0tsVYE"}},{"cell_type":"code","source":["y = df['heart disease'].values\n","y[0:5]"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"-3wWRpqJsXta","executionInfo":{"status":"ok","timestamp":1669557917198,"user_tz":-420,"elapsed":8,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"b34739d9-8e68-49dd-fb25-f3c9ce245b67"},"execution_count":39,"outputs":[{"output_type":"execute_result","data":{"text/plain":["array([1, 0, 1, 0, 0])"]},"metadata":{},"execution_count":39}]},{"cell_type":"code","source":["X = df.drop(columns=['heart disease'])\n","X"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":424},"id":"Bn2sNGbYsuaJ","executionInfo":{"status":"ok","timestamp":1669557917198,"user_tz":-420,"elapsed":7,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"359c53c1-62a7-4d2d-88a9-adff5f7c8c6d"},"execution_count":40,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" age sex BP cholestrol\n","0 70 1 130 322\n","1 67 0 115 564\n","2 57 1 124 261\n","3 64 1 128 263\n","4 74 0 120 269\n",".. ... ... ... ...\n","265 52 1 172 199\n","266 44 1 120 263\n","267 56 0 140 294\n","268 57 1 140 192\n","269 67 1 160 286\n","\n","[270 rows x 4 columns]"],"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"," \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"," \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"," \n","
agesexBPcholestrol
0701130322
1670115564
2571124261
3641128263
4740120269
...............
265521172199
266441120263
267560140294
268571140192
269671160286
\n","

270 rows × 4 columns

\n","
\n"," \n"," \n"," \n","\n"," \n","
\n","
\n"," "]},"metadata":{},"execution_count":40}]},{"cell_type":"markdown","source":["## Preprocessing Data (`Min-Max`)"],"metadata":{"id":"vVrv-KN8lRZd"}},{"cell_type":"code","source":["scaler = MinMaxScaler()\n","scaled = scaler.fit_transform(X)\n","features_names = X.columns.copy()\n","scaled_features = pd.DataFrame(scaled, columns=features_names)\n","scaled_features.head(10)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":363},"id":"CJG72qsluDIF","executionInfo":{"status":"ok","timestamp":1669557917199,"user_tz":-420,"elapsed":7,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"e311f141-8af0-4ab0-cf4e-c8fca61077d9"},"execution_count":41,"outputs":[{"output_type":"execute_result","data":{"text/plain":[" age sex BP cholestrol\n","0 0.854167 1.0 0.339623 0.447489\n","1 0.791667 0.0 0.198113 1.000000\n","2 0.583333 1.0 0.283019 0.308219\n","3 0.729167 1.0 0.320755 0.312785\n","4 0.937500 0.0 0.245283 0.326484\n","5 0.750000 1.0 0.245283 0.116438\n","6 0.562500 1.0 0.339623 0.296804\n","7 0.625000 1.0 0.150943 0.257991\n","8 0.645833 1.0 0.433962 0.381279\n","9 0.708333 0.0 0.528302 0.641553"],"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"," \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"," \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","
agesexBPcholestrol
00.8541671.00.3396230.447489
10.7916670.00.1981131.000000
20.5833331.00.2830190.308219
30.7291671.00.3207550.312785
40.9375000.00.2452830.326484
50.7500001.00.2452830.116438
60.5625001.00.3396230.296804
70.6250001.00.1509430.257991
80.6458331.00.4339620.381279
90.7083330.00.5283020.641553
\n","
\n"," \n"," \n"," \n","\n"," \n","
\n","
\n"," "]},"metadata":{},"execution_count":41}]},{"cell_type":"markdown","source":["### Split Data"],"metadata":{"id":"kGWWCZQ0vDU4"}},{"cell_type":"code","source":["from sklearn.model_selection import train_test_split\n","\n","X_train, X_test, y_train, y_test=train_test_split(scaled_features, y, test_size=0.2, random_state=1)\n"],"metadata":{"id":"SlLAZXahvAYm","executionInfo":{"status":"ok","timestamp":1669557917199,"user_tz":-420,"elapsed":7,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}}},"execution_count":42,"outputs":[]},{"cell_type":"markdown","source":["## Eksekusi Pada Model `Bagging Clasifier`"],"metadata":{"id":"Gf1redlcuafX"}},{"cell_type":"markdown","source":["### Bagging Clasifier Dengan SVC"],"metadata":{"id":"OxWXISa4pXc0"}},{"cell_type":"markdown","source":["Mencari akurasi tertinggi dengan N_estimators dari 2 sampai 100"],"metadata":{"id":"YAiGEWnupcN2"}},{"cell_type":"code","source":["# import model\n","from sklearn.naive_bayes import GaussianNB\n","from sklearn.svm import SVC\n","from sklearn.ensemble import BaggingClassifier\n","from sklearn.datasets import make_classification\n","from sklearn.metrics import accuracy_score\n","# eksekusi data pada model\n","X, y = make_classification(n_samples=100, n_features=4,\n"," n_informative=2, n_redundant=0,\n"," random_state=0, shuffle=False)\n","# bagging clasifier menggunakan SVC dan Gaussian(naive bayes)\n","# # SVC\n","n_estimator = range(2,101)\n","akurasi_bags_1 = []\n","for n in n_estimator:\n"," # inisialisasi model\n"," clf = BaggingClassifier(base_estimator=SVC(),\n"," n_estimators=n, random_state=40).fit(X_train, y_train)\n"," # predict x_test\n"," y_pred = clf.predict(X_test)\n"," # akurasi count\n"," akurasi_bags_1.append(accuracy_score(y_test,y_pred))\n","\n"," \n"],"metadata":{"id":"FmuXMRYGumNd","executionInfo":{"status":"ok","timestamp":1669557932212,"user_tz":-420,"elapsed":15019,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}}},"execution_count":43,"outputs":[]},{"cell_type":"markdown","source":["Visualisasi Akurasi Bagging dengan SVC"],"metadata":{"id":"NQz5QQcYrDmC"}},{"cell_type":"code","source":["import matplotlib.pyplot as plt\n","plt.plot(n_estimator,akurasi_bags_1)\n","plt.xlabel('Value of N')\n","plt.ylabel('Testing Accuracy')\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":0},"id":"L6HE-TdBrKhz","executionInfo":{"status":"ok","timestamp":1669557932213,"user_tz":-420,"elapsed":19,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"a761ff0d-639c-4bed-f1cf-f354286d5097"},"execution_count":44,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Mencari n_estimator dengan Akurasi Tertinggi"],"metadata":{"id":"6mb8XVulreQY"}},{"cell_type":"code","source":["akurasi_bags_1.index(max(akurasi_bags_1))+1 , max(akurasi_bags_1)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"63wpBduBrDJv","executionInfo":{"status":"ok","timestamp":1669557932213,"user_tz":-420,"elapsed":17,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"e2fb4835-1302-45a1-8d91-ff46234d9a51"},"execution_count":45,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(5, 0.7592592592592593)"]},"metadata":{},"execution_count":45}]},{"cell_type":"markdown","source":["### Bagging Clasifier Dengan GaussianNB"],"metadata":{"id":"LH0PLsNhuHik"}},{"cell_type":"markdown","source":["Mencari akurasi tertinggi dengan N_estimators dari 2 sampai 100"],"metadata":{"id":"YtwZWjBZubXW"}},{"cell_type":"code","source":["akurasi_bags_2= []\n","for n in n_estimator:\n"," # inisialisasi model\n"," clf2 = BaggingClassifier(base_estimator=GaussianNB(),\n"," n_estimators=n, random_state=40).fit(X_train, y_train)\n"," # predict x_test\n"," y_pred2 = clf2.predict(X_test)\n"," # akurasi count\n"," akurasi_bags_2.append(accuracy_score(y_test,y_pred2))"],"metadata":{"id":"qgfbA1GOuJqc","executionInfo":{"status":"ok","timestamp":1669557939863,"user_tz":-420,"elapsed":7664,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}}},"execution_count":46,"outputs":[]},{"cell_type":"code","source":["import joblib\n","clf2 = BaggingClassifier(base_estimator=GaussianNB(),\n"," n_estimators=6, random_state=40).fit(X_train, y_train)\n","filenameBCG = '/content/drive/MyDrive/datamining/tugas/model/bagginggaussian.pkl'\n","joblib.dump(clf2,filenameBCG)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"WfqxMOgPYj8g","executionInfo":{"status":"ok","timestamp":1669557939864,"user_tz":-420,"elapsed":19,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"ba8a28a2-c64e-44d6-a50e-ab7bbf7790c9"},"execution_count":47,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['/content/drive/MyDrive/datamining/tugas/model/bagginggaussian.pkl']"]},"metadata":{},"execution_count":47}]},{"cell_type":"markdown","source":["Visualisasi Hasil Akurasi "],"metadata":{"id":"u9Vlgo_eub2j"}},{"cell_type":"code","source":["plt.plot(n_estimator,akurasi_bags_2)\n","plt.xlabel('Value of N')\n","plt.ylabel('Testing Accuracy')\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":0},"id":"nmqKUXhVuXYy","executionInfo":{"status":"ok","timestamp":1669557939864,"user_tz":-420,"elapsed":17,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"c24e66b9-1c69-4bd7-9d92-2d58dacf8e0d"},"execution_count":48,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Mencari Akurasi Tertinggi"],"metadata":{"id":"MzaBZqEpudqq"}},{"cell_type":"code","source":["akurasi_bags_2.index(max(akurasi_bags_2))+1 , max(akurasi_bags_2)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"PYj_oNYkugfq","executionInfo":{"status":"ok","timestamp":1669557939865,"user_tz":-420,"elapsed":17,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"04a22fbc-f0ec-46d2-df64-9336f5a53e62"},"execution_count":49,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(7, 0.7777777777777778)"]},"metadata":{},"execution_count":49}]},{"cell_type":"markdown","source":["## Eksekusi Pada Model `Random Forest`"],"metadata":{"id":"sxR5a0x-wUqZ"}},{"cell_type":"markdown","source":["Mencari akurasi tertinggi dengan N_estimators dari 2 sampai 100"],"metadata":{"id":"Xo10W7C7xwp7"}},{"cell_type":"code","source":["from sklearn.ensemble import RandomForestClassifier\n","akurasirf= []\n","for n in n_estimator:\n"," # inisialisasi model\n"," rf = RandomForestClassifier(\n"," n_estimators=n,max_depth=2, random_state=40).fit(X_train, y_train)\n"," # predict x_test\n"," y_predrf = rf.predict(X_test)\n"," # akurasi count\n"," akurasirf.append(accuracy_score(y_test,y_predrf))"],"metadata":{"id":"0DXRtRnkxxFN","executionInfo":{"status":"ok","timestamp":1669557947698,"user_tz":-420,"elapsed":7848,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}}},"execution_count":50,"outputs":[]},{"cell_type":"code","source":["rf = RandomForestClassifier(\n"," n_estimators=13,max_depth=2, random_state=40).fit(X_train, y_train)\n","filenameRF = '/content/drive/MyDrive/datamining/tugas/model/randomforest.pkl'\n","joblib.dump(rf,filenameRF)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"xgbKAWqtZLXN","executionInfo":{"status":"ok","timestamp":1669557947701,"user_tz":-420,"elapsed":22,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"48de2222-f17d-4ea6-c248-527c526ec43e"},"execution_count":51,"outputs":[{"output_type":"execute_result","data":{"text/plain":["['/content/drive/MyDrive/datamining/tugas/model/randomforest.pkl']"]},"metadata":{},"execution_count":51}]},{"cell_type":"markdown","source":["Visualisasi Hasil Akurasi "],"metadata":{"id":"cvvBwCB1yOG7"}},{"cell_type":"code","source":["plt.plot(n_estimator,akurasirf)\n","plt.xlabel('Value of N')\n","plt.ylabel('Testing Accuracy')\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":0},"id":"sJBi98r0ySH6","executionInfo":{"status":"ok","timestamp":1669557947702,"user_tz":-420,"elapsed":21,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"d48686e6-0175-4864-d4f4-ada6bd3edc76"},"execution_count":52,"outputs":[{"output_type":"display_data","data":{"text/plain":["
"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":["Mencari Akurasi Tertinggi dari N"],"metadata":{"id":"QPwR_ERIyR3a"}},{"cell_type":"code","source":["akurasirf.index(max(akurasirf))+1 , max(akurasirf)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"M6-f_7ZsyYyN","executionInfo":{"status":"ok","timestamp":1669557947703,"user_tz":-420,"elapsed":21,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"0c9ac547-4ced-460d-c2a0-e00b3683e5c2"},"execution_count":53,"outputs":[{"output_type":"execute_result","data":{"text/plain":["(14, 0.7222222222222222)"]},"metadata":{},"execution_count":53}]},{"cell_type":"markdown","source":["## Eksekusi Pada Model `Stacking clasifier`"],"metadata":{"id":"j-G1UqBkyzje"}},{"cell_type":"code","source":["from sklearn.ensemble import StackingClassifier\n","from sklearn.tree import DecisionTreeClassifier\n","\n","# estimator menggunakan Random Forest, SVC GaussianNB\n","## untuk n_estimators menggunakan n dengan akurasi tertinggi \n","estimators = [\n"," ('rf', RandomForestClassifier(n_estimators=38, max_depth=2, random_state=40)),\n"," ('svc', SVC()),\n"," ('gnb', GaussianNB()),\n"," ('bagsvc', BaggingClassifier(base_estimator=SVC(),\n"," n_estimators=14, random_state=40)),\n"," ('baggnb' ,BaggingClassifier(base_estimator=GaussianNB(),\n"," n_estimators=9, random_state=40))\n","]\n","\n","sc = StackingClassifier(\n"," estimators=estimators, final_estimator=SVC()).fit(X_train, y_train)\n","\n","y_predsc = sc.predict(X_test)\n","akurasi = accuracy_score(y_test,y_predsc)"],"metadata":{"id":"xNsAZ2Ag1qT-","executionInfo":{"status":"ok","timestamp":1669557948289,"user_tz":-420,"elapsed":604,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}}},"execution_count":54,"outputs":[]},{"cell_type":"markdown","source":["### Hasil Akurasi Dan Score dari Stacking Clasifier"],"metadata":{"id":"QAkUH2ex38_T"}},{"cell_type":"code","source":["print(f'Akurasi Untuk Stacking Clasifier = {akurasi}')"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"nuq9t-_d4DP5","executionInfo":{"status":"ok","timestamp":1669557948289,"user_tz":-420,"elapsed":4,"user":{"displayName":"Caca Erha","userId":"13359221303846732984"}},"outputId":"030f2372-8398-44cb-a05a-44fd8223383e"},"execution_count":55,"outputs":[{"output_type":"stream","name":"stdout","text":["Akurasi Untuk Stacking Clasifier = 0.6666666666666666\n"]}]}]}