neural-network-keras-tuner.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":65862,"databundleVersionId":7469115,"sourceType":"competition"},{"sourceId":7637222,"sourceType":"datasetVersion","datasetId":4450702}],"dockerImageVersionId":30646,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2024-02-16T21:01:37.592465Z","iopub.execute_input":"2024-02-16T21:01:37.593195Z","iopub.status.idle":"2024-02-16T21:01:39.016024Z","shell.execute_reply.started":"2024-02-16T21:01:37.593124Z","shell.execute_reply":"2024-02-16T21:01:39.014828Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stdout","text":"/kaggle/input/widsdatathon2024-challenge1/sample_submission.csv\n/kaggle/input/widsdatathon2024-challenge1/training.csv\n/kaggle/input/widsdatathon2024-challenge1/test.csv\n/kaggle/input/equity-in-healthcare-clean-datasets/new_test (1).csv\n/kaggle/input/equity-in-healthcare-clean-datasets/new_train (1).csv\n","output_type":"stream"}]},{"cell_type":"code","source":"test = pd.read_csv('/kaggle/input/widsdatathon2024-challenge1/test.csv')\ntrain = pd.read_csv('/kaggle/input/widsdatathon2024-challenge1/training.csv')","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:39.018090Z","iopub.execute_input":"2024-02-16T21:01:39.018564Z","iopub.status.idle":"2024-02-16T21:01:39.695974Z","shell.execute_reply.started":"2024-02-16T21:01:39.018533Z","shell.execute_reply":"2024-02-16T21:01:39.694727Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"code","source":"new_train = pd.read_csv('/kaggle/input/equity-in-healthcare-clean-datasets/new_train (1).csv')\nnew_test = pd.read_csv('/kaggle/input/equity-in-healthcare-clean-datasets/new_test (1).csv')","metadata":{"_kg_hide-input":true,"execution":{"iopub.status.busy":"2024-02-16T21:01:39.697537Z","iopub.execute_input":"2024-02-16T21:01:39.697977Z","iopub.status.idle":"2024-02-16T21:01:40.344910Z","shell.execute_reply.started":"2024-02-16T21:01:39.697937Z","shell.execute_reply":"2024-02-16T21:01:40.343626Z"},"trusted":true},"execution_count":3,"outputs":[]},{"cell_type":"code","source":"new_train.head()","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.347076Z","iopub.execute_input":"2024-02-16T21:01:40.347684Z","iopub.status.idle":"2024-02-16T21:01:40.397391Z","shell.execute_reply.started":"2024-02-16T21:01:40.347642Z","shell.execute_reply":"2024-02-16T21:01:40.396021Z"},"trusted":true},"execution_count":4,"outputs":[{"execution_count":4,"output_type":"execute_result","data":{"text/plain":"   index  patient_id patient_state  patient_zip3  patient_age  \\\n0      0      475714            CA           924           84   \n1      1      349367            CA           928           62   \n2      2      138632            TX           760           43   \n3      3      617843            CA           926           45   \n4      4      817482            ID           836           55   \n\n  breast_cancer_diagnosis_code  \\\n0                       C50919   \n1                       C50411   \n2                       C50112   \n3                       C50212   \n4                         1749   \n\n                        breast_cancer_diagnosis_desc  \\\n0  Malignant neoplasm of unsp site of unspecified...   \n1  Malig neoplm of upper-outer quadrant of right ...   \n2  Malignant neoplasm of central portion of left ...   \n3  Malig neoplasm of upper-inner quadrant of left...   \n4  Malignant neoplasm of breast (female), unspeci...   \n\n  metastatic_cancer_diagnosis_code Region            Division  ...  young_ind  \\\n0                            C7989   West             Pacific  ...          0   \n1                             C773   West             Pacific  ...          0   \n2                             C773  South  West South Central  ...          0   \n3                             C773   West             Pacific  ...          0   \n4                             C773   West            Mountain  ...          0   \n\n   desolated  home_wealth  air_quality  wealth_index  education_ratio  \\\n0   0.023382    11.798909    79.494293     21.602549         2.863469   \n1   0.013331    41.653200    70.901475      2.917221         0.395453   \n2   0.050133    17.292436    62.590311      2.228523         0.362889   \n3   0.014535    59.960170    65.193591      1.063895         0.104685   \n4   0.223895    19.262605    57.189004      1.124218         0.588750   \n\n   hh_income_ratio  age_ratio  race_ration  DiagPeriodL90D  \n0         0.373972   4.444668     2.779721               1  \n1         0.078278   2.413623     1.553169               1  \n2         0.095774   2.649374     0.574321               1  \n3         0.082758   1.588927     0.795576               0  \n4         0.204895   1.780586     0.307016               0  \n\n[5 rows x 92 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>index</th>\n      <th>patient_id</th>\n      <th>patient_state</th>\n      <th>patient_zip3</th>\n      <th>patient_age</th>\n      <th>breast_cancer_diagnosis_code</th>\n      <th>breast_cancer_diagnosis_desc</th>\n      <th>metastatic_cancer_diagnosis_code</th>\n      <th>Region</th>\n      <th>Division</th>\n      <th>...</th>\n      <th>young_ind</th>\n      <th>desolated</th>\n      <th>home_wealth</th>\n      <th>air_quality</th>\n      <th>wealth_index</th>\n      <th>education_ratio</th>\n      <th>hh_income_ratio</th>\n      <th>age_ratio</th>\n      <th>race_ration</th>\n      <th>DiagPeriodL90D</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>475714</td>\n      <td>CA</td>\n      <td>924</td>\n      <td>84</td>\n      <td>C50919</td>\n      <td>Malignant neoplasm of unsp site of unspecified...</td>\n      <td>C7989</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0.023382</td>\n      <td>11.798909</td>\n      <td>79.494293</td>\n      <td>21.602549</td>\n      <td>2.863469</td>\n      <td>0.373972</td>\n      <td>4.444668</td>\n      <td>2.779721</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>349367</td>\n      <td>CA</td>\n      <td>928</td>\n      <td>62</td>\n      <td>C50411</td>\n      <td>Malig neoplm of upper-outer quadrant of right ...</td>\n      <td>C773</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0.013331</td>\n      <td>41.653200</td>\n      <td>70.901475</td>\n      <td>2.917221</td>\n      <td>0.395453</td>\n      <td>0.078278</td>\n      <td>2.413623</td>\n      <td>1.553169</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>138632</td>\n      <td>TX</td>\n      <td>760</td>\n      <td>43</td>\n      <td>C50112</td>\n      <td>Malignant neoplasm of central portion of left ...</td>\n      <td>C773</td>\n      <td>South</td>\n      <td>West South Central</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0.050133</td>\n      <td>17.292436</td>\n      <td>62.590311</td>\n      <td>2.228523</td>\n      <td>0.362889</td>\n      <td>0.095774</td>\n      <td>2.649374</td>\n      <td>0.574321</td>\n      <td>1</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>617843</td>\n      <td>CA</td>\n      <td>926</td>\n      <td>45</td>\n      <td>C50212</td>\n      <td>Malig neoplasm of upper-inner quadrant of left...</td>\n      <td>C773</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0.014535</td>\n      <td>59.960170</td>\n      <td>65.193591</td>\n      <td>1.063895</td>\n      <td>0.104685</td>\n      <td>0.082758</td>\n      <td>1.588927</td>\n      <td>0.795576</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>817482</td>\n      <td>ID</td>\n      <td>836</td>\n      <td>55</td>\n      <td>1749</td>\n      <td>Malignant neoplasm of breast (female), unspeci...</td>\n      <td>C773</td>\n      <td>West</td>\n      <td>Mountain</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0.223895</td>\n      <td>19.262605</td>\n      <td>57.189004</td>\n      <td>1.124218</td>\n      <td>0.588750</td>\n      <td>0.204895</td>\n      <td>1.780586</td>\n      <td>0.307016</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 92 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"new_test.head()","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.399234Z","iopub.execute_input":"2024-02-16T21:01:40.399916Z","iopub.status.idle":"2024-02-16T21:01:40.431673Z","shell.execute_reply.started":"2024-02-16T21:01:40.399871Z","shell.execute_reply":"2024-02-16T21:01:40.430505Z"},"trusted":true},"execution_count":5,"outputs":[{"execution_count":5,"output_type":"execute_result","data":{"text/plain":"   index  patient_id patient_state  patient_zip3  patient_age  \\\n0      0      573710            IN           467           54   \n1      1      593679            FL           337           52   \n2      2      184532            CA           917           61   \n3      3      447383            CA           917           64   \n4      4      687972            CA           900           40   \n\n  breast_cancer_diagnosis_code  \\\n0                       C50412   \n1                       C50912   \n2                       C50911   \n3                       C50912   \n4                       C50412   \n\n                        breast_cancer_diagnosis_desc  \\\n0  Malig neoplasm of upper-outer quadrant of left...   \n1  Malignant neoplasm of unspecified site of left...   \n2  Malignant neoplasm of unsp site of right femal...   \n3  Malignant neoplasm of unspecified site of left...   \n4  Malig neoplasm of upper-outer quadrant of left...   \n\n  metastatic_cancer_diagnosis_code   Region            Division  ...  obese  \\\n0                             C773  Midwest  East North Central  ...      1   \n1                             C787    South      South Atlantic  ...      0   \n2                             C773     West             Pacific  ...      1   \n3                             C779     West             Pacific  ...      1   \n4                             C779     West             Pacific  ...      0   \n\n   young_ind  desolated  home_wealth  air_quality  wealth_index  \\\n0          0   0.291463    12.848207    57.829202      0.285810   \n1          0   0.015403    18.257627    58.357215      2.042298   \n2          0   0.015880    35.156794    76.990275      6.033070   \n3          0   0.015880    35.156794    76.990275      6.033070   \n4          0   0.005800    27.527409    73.998150     27.473246   \n\n   education_ratio  hh_income_ratio  age_ratio  race_ration  \n0         0.692736         0.123927   2.103073     0.092442  \n1         0.240347         0.203070   0.847931     0.409605  \n2         0.571530         0.090249   2.177161     2.815961  \n3         0.571530         0.090249   2.177161     2.815961  \n4         0.609851         0.283686   2.003077     2.398118  \n\n[5 rows x 91 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>index</th>\n      <th>patient_id</th>\n      <th>patient_state</th>\n      <th>patient_zip3</th>\n      <th>patient_age</th>\n      <th>breast_cancer_diagnosis_code</th>\n      <th>breast_cancer_diagnosis_desc</th>\n      <th>metastatic_cancer_diagnosis_code</th>\n      <th>Region</th>\n      <th>Division</th>\n      <th>...</th>\n      <th>obese</th>\n      <th>young_ind</th>\n      <th>desolated</th>\n      <th>home_wealth</th>\n      <th>air_quality</th>\n      <th>wealth_index</th>\n      <th>education_ratio</th>\n      <th>hh_income_ratio</th>\n      <th>age_ratio</th>\n      <th>race_ration</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>573710</td>\n      <td>IN</td>\n      <td>467</td>\n      <td>54</td>\n      <td>C50412</td>\n      <td>Malig neoplasm of upper-outer quadrant of left...</td>\n      <td>C773</td>\n      <td>Midwest</td>\n      <td>East North Central</td>\n      <td>...</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0.291463</td>\n      <td>12.848207</td>\n      <td>57.829202</td>\n      <td>0.285810</td>\n      <td>0.692736</td>\n      <td>0.123927</td>\n      <td>2.103073</td>\n      <td>0.092442</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>593679</td>\n      <td>FL</td>\n      <td>337</td>\n      <td>52</td>\n      <td>C50912</td>\n      <td>Malignant neoplasm of unspecified site of left...</td>\n      <td>C787</td>\n      <td>South</td>\n      <td>South Atlantic</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0.015403</td>\n      <td>18.257627</td>\n      <td>58.357215</td>\n      <td>2.042298</td>\n      <td>0.240347</td>\n      <td>0.203070</td>\n      <td>0.847931</td>\n      <td>0.409605</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>184532</td>\n      <td>CA</td>\n      <td>917</td>\n      <td>61</td>\n      <td>C50911</td>\n      <td>Malignant neoplasm of unsp site of right femal...</td>\n      <td>C773</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0.015880</td>\n      <td>35.156794</td>\n      <td>76.990275</td>\n      <td>6.033070</td>\n      <td>0.571530</td>\n      <td>0.090249</td>\n      <td>2.177161</td>\n      <td>2.815961</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>447383</td>\n      <td>CA</td>\n      <td>917</td>\n      <td>64</td>\n      <td>C50912</td>\n      <td>Malignant neoplasm of unspecified site of left...</td>\n      <td>C779</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0.015880</td>\n      <td>35.156794</td>\n      <td>76.990275</td>\n      <td>6.033070</td>\n      <td>0.571530</td>\n      <td>0.090249</td>\n      <td>2.177161</td>\n      <td>2.815961</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>687972</td>\n      <td>CA</td>\n      <td>900</td>\n      <td>40</td>\n      <td>C50412</td>\n      <td>Malig neoplasm of upper-outer quadrant of left...</td>\n      <td>C779</td>\n      <td>West</td>\n      <td>Pacific</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0.005800</td>\n      <td>27.527409</td>\n      <td>73.998150</td>\n      <td>27.473246</td>\n      <td>0.609851</td>\n      <td>0.283686</td>\n      <td>2.003077</td>\n      <td>2.398118</td>\n    </tr>\n  </tbody>\n</table>\n<p>5 rows × 91 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"new_test.dtypes","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.435090Z","iopub.execute_input":"2024-02-16T21:01:40.436641Z","iopub.status.idle":"2024-02-16T21:01:40.447042Z","shell.execute_reply.started":"2024-02-16T21:01:40.436589Z","shell.execute_reply":"2024-02-16T21:01:40.445500Z"},"trusted":true},"execution_count":6,"outputs":[{"execution_count":6,"output_type":"execute_result","data":{"text/plain":"index                int64\npatient_id           int64\npatient_state       object\npatient_zip3         int64\npatient_age          int64\n                    ...   \nwealth_index       float64\neducation_ratio    float64\nhh_income_ratio    float64\nage_ratio          float64\nrace_ration        float64\nLength: 91, dtype: object"},"metadata":{}}]},{"cell_type":"code","source":"#values = {'patient_race' : 'missing', 'payer_type': 'missing'}\n#train.fillna(value=values, inplace = True)\n\n#if it works do the same for test\n#test.fillna(value=values, inplace = True)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.448802Z","iopub.execute_input":"2024-02-16T21:01:40.449258Z","iopub.status.idle":"2024-02-16T21:01:40.463737Z","shell.execute_reply.started":"2024-02-16T21:01:40.449225Z","shell.execute_reply":"2024-02-16T21:01:40.462107Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"markdown","source":"# I correct some datatypes to category dtype","metadata":{}},{"cell_type":"code","source":"types_dict = new_train.dtypes.to_dict()\ntest_dtypes = new_test.dtypes.to_dict()","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.465593Z","iopub.execute_input":"2024-02-16T21:01:40.466474Z","iopub.status.idle":"2024-02-16T21:01:40.485074Z","shell.execute_reply.started":"2024-02-16T21:01:40.466382Z","shell.execute_reply":"2024-02-16T21:01:40.483538Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"code","source":"for key in types_dict:\n    if types_dict[key]=='O':\n        new_train[key] = new_train[key].astype('category')","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.486813Z","iopub.execute_input":"2024-02-16T21:01:40.487251Z","iopub.status.idle":"2024-02-16T21:01:40.527196Z","shell.execute_reply.started":"2024-02-16T21:01:40.487211Z","shell.execute_reply":"2024-02-16T21:01:40.525982Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"code","source":"for key in test_dtypes:\n    if test_dtypes[key]=='O':\n        new_test[key] = new_test[key].astype('category')","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.531473Z","iopub.execute_input":"2024-02-16T21:01:40.532224Z","iopub.status.idle":"2024-02-16T21:01:40.554320Z","shell.execute_reply.started":"2024-02-16T21:01:40.532181Z","shell.execute_reply":"2024-02-16T21:01:40.552265Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"code","source":"#changing patient_zip3 into categorical variable\nnew_train['patient_zip3'] = new_train['patient_zip3'].astype('category')\nnew_test['patient_zip3'] = new_test['patient_zip3'].astype('category')\nnew_train['DiagPeriodL90D']=train['DiagPeriodL90D'].astype('category')","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.556517Z","iopub.execute_input":"2024-02-16T21:01:40.556931Z","iopub.status.idle":"2024-02-16T21:01:40.566464Z","shell.execute_reply.started":"2024-02-16T21:01:40.556892Z","shell.execute_reply":"2024-02-16T21:01:40.565023Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"code","source":"","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.567995Z","iopub.execute_input":"2024-02-16T21:01:40.568454Z","iopub.status.idle":"2024-02-16T21:01:40.590693Z","shell.execute_reply.started":"2024-02-16T21:01:40.568411Z","shell.execute_reply":"2024-02-16T21:01:40.589723Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"new_test.columns","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.592412Z","iopub.execute_input":"2024-02-16T21:01:40.593977Z","iopub.status.idle":"2024-02-16T21:01:40.608108Z","shell.execute_reply.started":"2024-02-16T21:01:40.593932Z","shell.execute_reply":"2024-02-16T21:01:40.606714Z"},"trusted":true},"execution_count":13,"outputs":[{"execution_count":13,"output_type":"execute_result","data":{"text/plain":"Index(['index', 'patient_id', 'patient_state', 'patient_zip3', 'patient_age',\n       'breast_cancer_diagnosis_code', 'breast_cancer_diagnosis_desc',\n       'metastatic_cancer_diagnosis_code', 'Region', 'Division', 'population',\n       'density', 'age_median', 'age_under_10', 'age_10_to_19', 'age_20s',\n       'age_30s', 'age_40s', 'age_50s', 'age_60s', 'age_70s', 'age_over_80',\n       'male', 'female', 'married', 'divorced', 'never_married', 'widowed',\n       'family_size', 'family_dual_income', 'income_household_median',\n       'income_household_under_5', 'income_household_5_to_10',\n       'income_household_10_to_15', 'income_household_15_to_20',\n       'income_household_20_to_25', 'income_household_25_to_35',\n       'income_household_35_to_50', 'income_household_50_to_75',\n       'income_household_75_to_100', 'income_household_100_to_150',\n       'income_household_150_over', 'income_household_six_figure',\n       'income_individual_median', 'home_ownership', 'housing_units',\n       'home_value', 'rent_median', 'rent_burden', 'education_less_highschool',\n       'education_highschool', 'education_some_college', 'education_bachelors',\n       'education_graduate', 'education_college_or_above',\n       'education_stem_degree', 'labor_force_participation',\n       'unemployment_rate', 'self_employed', 'farmer', 'race_white',\n       'race_black', 'race_asian', 'race_native', 'race_pacific', 'race_other',\n       'race_multiple', 'hispanic', 'disabled', 'poverty', 'limited_english',\n       'commute_time', 'health_uninsured', 'veteran', 'Ozone', 'PM25', 'N02',\n       'payer_type', 'patient_race', 'bmi', 'underweight', 'obese',\n       'young_ind', 'desolated', 'home_wealth', 'air_quality', 'wealth_index',\n       'education_ratio', 'hh_income_ratio', 'age_ratio', 'race_ration'],\n      dtype='object')"},"metadata":{}}]},{"cell_type":"code","source":"new_train.columns","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.609942Z","iopub.execute_input":"2024-02-16T21:01:40.610613Z","iopub.status.idle":"2024-02-16T21:01:40.627399Z","shell.execute_reply.started":"2024-02-16T21:01:40.610566Z","shell.execute_reply":"2024-02-16T21:01:40.626229Z"},"trusted":true},"execution_count":14,"outputs":[{"execution_count":14,"output_type":"execute_result","data":{"text/plain":"Index(['index', 'patient_id', 'patient_state', 'patient_zip3', 'patient_age',\n       'breast_cancer_diagnosis_code', 'breast_cancer_diagnosis_desc',\n       'metastatic_cancer_diagnosis_code', 'Region', 'Division', 'population',\n       'density', 'age_median', 'age_under_10', 'age_10_to_19', 'age_20s',\n       'age_30s', 'age_40s', 'age_50s', 'age_60s', 'age_70s', 'age_over_80',\n       'male', 'female', 'married', 'divorced', 'never_married', 'widowed',\n       'family_size', 'family_dual_income', 'income_household_median',\n       'income_household_under_5', 'income_household_5_to_10',\n       'income_household_10_to_15', 'income_household_15_to_20',\n       'income_household_20_to_25', 'income_household_25_to_35',\n       'income_household_35_to_50', 'income_household_50_to_75',\n       'income_household_75_to_100', 'income_household_100_to_150',\n       'income_household_150_over', 'income_household_six_figure',\n       'income_individual_median', 'home_ownership', 'housing_units',\n       'home_value', 'rent_median', 'rent_burden', 'education_less_highschool',\n       'education_highschool', 'education_some_college', 'education_bachelors',\n       'education_graduate', 'education_college_or_above',\n       'education_stem_degree', 'labor_force_participation',\n       'unemployment_rate', 'self_employed', 'farmer', 'race_white',\n       'race_black', 'race_asian', 'race_native', 'race_pacific', 'race_other',\n       'race_multiple', 'hispanic', 'disabled', 'poverty', 'limited_english',\n       'commute_time', 'health_uninsured', 'veteran', 'Ozone', 'PM25', 'N02',\n       'payer_type', 'patient_race', 'bmi', 'underweight', 'obese',\n       'young_ind', 'desolated', 'home_wealth', 'air_quality', 'wealth_index',\n       'education_ratio', 'hh_income_ratio', 'age_ratio', 'race_ration',\n       'DiagPeriodL90D'],\n      dtype='object')"},"metadata":{}}]},{"cell_type":"code","source":"new_train.drop(columns = ['index', 'patient_id'], inplace= True)\nnew_test.drop(columns = ['index', 'patient_id'], inplace= True)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.628951Z","iopub.execute_input":"2024-02-16T21:01:40.629655Z","iopub.status.idle":"2024-02-16T21:01:40.649080Z","shell.execute_reply.started":"2024-02-16T21:01:40.629610Z","shell.execute_reply":"2024-02-16T21:01:40.648212Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"num_col = []\ncat_col = []\nfor col in new_train.columns:\n    if new_train[col].dtypes in ['float64' , 'int64']:\n        num_col.append(col)\n    else: cat_col.append(col)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.651670Z","iopub.execute_input":"2024-02-16T21:01:40.652143Z","iopub.status.idle":"2024-02-16T21:01:40.668486Z","shell.execute_reply.started":"2024-02-16T21:01:40.652083Z","shell.execute_reply":"2024-02-16T21:01:40.667328Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"print(num_col)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.670249Z","iopub.execute_input":"2024-02-16T21:01:40.670890Z","iopub.status.idle":"2024-02-16T21:01:40.688233Z","shell.execute_reply.started":"2024-02-16T21:01:40.670843Z","shell.execute_reply":"2024-02-16T21:01:40.686556Z"},"trusted":true},"execution_count":17,"outputs":[{"name":"stdout","text":"['patient_age', 'population', 'density', 'age_median', 'age_under_10', 'age_10_to_19', 'age_20s', 'age_30s', 'age_40s', 'age_50s', 'age_60s', 'age_70s', 'age_over_80', 'male', 'female', 'married', 'divorced', 'never_married', 'widowed', 'family_size', 'family_dual_income', 'income_household_median', 'income_household_under_5', 'income_household_5_to_10', 'income_household_10_to_15', 'income_household_15_to_20', 'income_household_20_to_25', 'income_household_25_to_35', 'income_household_35_to_50', 'income_household_50_to_75', 'income_household_75_to_100', 'income_household_100_to_150', 'income_household_150_over', 'income_household_six_figure', 'income_individual_median', 'home_ownership', 'housing_units', 'home_value', 'rent_median', 'rent_burden', 'education_less_highschool', 'education_highschool', 'education_some_college', 'education_bachelors', 'education_graduate', 'education_college_or_above', 'education_stem_degree', 'labor_force_participation', 'unemployment_rate', 'self_employed', 'farmer', 'race_white', 'race_black', 'race_asian', 'race_native', 'race_pacific', 'race_other', 'race_multiple', 'hispanic', 'disabled', 'poverty', 'limited_english', 'commute_time', 'health_uninsured', 'veteran', 'Ozone', 'PM25', 'N02', 'bmi', 'underweight', 'obese', 'young_ind', 'desolated', 'home_wealth', 'air_quality', 'wealth_index', 'education_ratio', 'hh_income_ratio', 'age_ratio', 'race_ration']\n","output_type":"stream"}]},{"cell_type":"code","source":"cat_col = ['patient_state', 'patient_zip3', 'breast_cancer_diagnosis_code', 'breast_cancer_diagnosis_desc', 'metastatic_cancer_diagnosis_code', 'Region', 'Division', 'payer_type', 'patient_race']","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.691018Z","iopub.execute_input":"2024-02-16T21:01:40.691465Z","iopub.status.idle":"2024-02-16T21:01:40.707973Z","shell.execute_reply.started":"2024-02-16T21:01:40.691434Z","shell.execute_reply":"2024-02-16T21:01:40.706489Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"markdown","source":"# I split my columns into numeric and categorical so I can feed them into a pipeline and transform numerical values with the Standard Scaler and categorical data with OneHotEncoder","metadata":{}},{"cell_type":"code","source":"from sklearn.model_selection import train_test_split\nfrom sklearn.preprocessing import StandardScaler, OneHotEncoder\nfrom sklearn.compose import make_column_transformer\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.pipeline import make_pipeline","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:40.710112Z","iopub.execute_input":"2024-02-16T21:01:40.710477Z","iopub.status.idle":"2024-02-16T21:01:42.735349Z","shell.execute_reply.started":"2024-02-16T21:01:40.710446Z","shell.execute_reply":"2024-02-16T21:01:42.734263Z"},"trusted":true},"execution_count":19,"outputs":[]},{"cell_type":"markdown","source":"# I create my pipeline","metadata":{}},{"cell_type":"code","source":"preprocessor = make_column_transformer(\n    (make_pipeline(StandardScaler()), num_col),\n    (OneHotEncoder(), cat_col)\n)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.736906Z","iopub.execute_input":"2024-02-16T21:01:42.737746Z","iopub.status.idle":"2024-02-16T21:01:42.744685Z","shell.execute_reply.started":"2024-02-16T21:01:42.737698Z","shell.execute_reply":"2024-02-16T21:01:42.743193Z"},"trusted":true},"execution_count":20,"outputs":[]},{"cell_type":"code","source":"y = train['DiagPeriodL90D']","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.749116Z","iopub.execute_input":"2024-02-16T21:01:42.749498Z","iopub.status.idle":"2024-02-16T21:01:42.759275Z","shell.execute_reply.started":"2024-02-16T21:01:42.749470Z","shell.execute_reply":"2024-02-16T21:01:42.758123Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"code","source":"train_X = new_train.drop(columns=['DiagPeriodL90D'])\ntest_X = new_test\nprint( train_X.shape, test_X.shape)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.760757Z","iopub.execute_input":"2024-02-16T21:01:42.761229Z","iopub.status.idle":"2024-02-16T21:01:42.780944Z","shell.execute_reply.started":"2024-02-16T21:01:42.761188Z","shell.execute_reply":"2024-02-16T21:01:42.779197Z"},"trusted":true},"execution_count":22,"outputs":[{"name":"stdout","text":"(12906, 89) (5792, 89)\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# I create one big dataframe","metadata":{}},{"cell_type":"code","source":"whole_df = pd.concat([train_X, test_X], axis =0)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.783099Z","iopub.execute_input":"2024-02-16T21:01:42.784404Z","iopub.status.idle":"2024-02-16T21:01:42.807654Z","shell.execute_reply.started":"2024-02-16T21:01:42.784354Z","shell.execute_reply":"2024-02-16T21:01:42.806423Z"},"trusted":true},"execution_count":23,"outputs":[]},{"cell_type":"code","source":"#train_X_pp = preprocessor.fit_transform(train_X)\n#test_X_pp = preprocessor.fit_transform(test_X)\n#print(train_X_pp.shape, test_X_pp.shape)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.812031Z","iopub.execute_input":"2024-02-16T21:01:42.812691Z","iopub.status.idle":"2024-02-16T21:01:42.817585Z","shell.execute_reply.started":"2024-02-16T21:01:42.812655Z","shell.execute_reply":"2024-02-16T21:01:42.816464Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"code","source":"whole_sparse = preprocessor.fit_transform(whole_df)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:42.819302Z","iopub.execute_input":"2024-02-16T21:01:42.819994Z","iopub.status.idle":"2024-02-16T21:01:43.104370Z","shell.execute_reply.started":"2024-02-16T21:01:42.819952Z","shell.execute_reply":"2024-02-16T21:01:43.102836Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# I change my data from sparce format into a dense format","metadata":{}},{"cell_type":"code","source":"whole_dense = whole_sparse.toarray()","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.105778Z","iopub.execute_input":"2024-02-16T21:01:43.106303Z","iopub.status.idle":"2024-02-16T21:01:43.356878Z","shell.execute_reply.started":"2024-02-16T21:01:43.106173Z","shell.execute_reply":"2024-02-16T21:01:43.355857Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"code","source":"whole_dense.shape","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.358354Z","iopub.execute_input":"2024-02-16T21:01:43.359425Z","iopub.status.idle":"2024-02-16T21:01:43.368039Z","shell.execute_reply.started":"2024-02-16T21:01:43.359378Z","shell.execute_reply":"2024-02-16T21:01:43.366631Z"},"trusted":true},"execution_count":27,"outputs":[{"execution_count":27,"output_type":"execute_result","data":{"text/plain":"(18698, 1060)"},"metadata":{}}]},{"cell_type":"markdown","source":"# And split back the data into test and train sets","metadata":{}},{"cell_type":"code","source":"train_X_dense = whole_dense[:12906, :]\n\ntest_X_dense = whole_dense[12906:, :]","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.375875Z","iopub.execute_input":"2024-02-16T21:01:43.376272Z","iopub.status.idle":"2024-02-16T21:01:43.381886Z","shell.execute_reply.started":"2024-02-16T21:01:43.376243Z","shell.execute_reply":"2024-02-16T21:01:43.380667Z"},"trusted":true},"execution_count":28,"outputs":[]},{"cell_type":"code","source":"print(train_X_dense.shape, test_X_dense.shape)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.384066Z","iopub.execute_input":"2024-02-16T21:01:43.385055Z","iopub.status.idle":"2024-02-16T21:01:43.393910Z","shell.execute_reply.started":"2024-02-16T21:01:43.385004Z","shell.execute_reply":"2024-02-16T21:01:43.392940Z"},"trusted":true},"execution_count":29,"outputs":[{"name":"stdout","text":"(12906, 1060) (5792, 1060)\n","output_type":"stream"}]},{"cell_type":"code","source":"print(train_X_dense)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.395684Z","iopub.execute_input":"2024-02-16T21:01:43.396398Z","iopub.status.idle":"2024-02-16T21:01:43.407879Z","shell.execute_reply.started":"2024-02-16T21:01:43.396354Z","shell.execute_reply":"2024-02-16T21:01:43.406472Z"},"trusted":true},"execution_count":30,"outputs":[{"name":"stdout","text":"[[ 1.87012152  0.78398917 -0.12589829 ...  0.          0.\n   1.        ]\n [ 0.21036103  1.33966983  0.25031455 ...  0.          0.\n   1.        ]\n [-1.22306848  0.10125487 -0.31745184 ...  0.          0.\n   1.        ]\n ...\n [ 0.13491737  0.88216276  0.11439412 ...  0.          0.\n   1.        ]\n [-1.67573043  3.67200364  5.36128109 ...  0.          0.\n   0.        ]\n [-2.2792797   0.6948772  -0.01449712 ...  1.          0.\n   0.        ]]\n","output_type":"stream"}]},{"cell_type":"code","source":"X_train, X_test, y_train, y_test = train_test_split(train_X_dense, y, test_size=0.2, stratify=y, shuffle=True, random_state=66)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.409539Z","iopub.execute_input":"2024-02-16T21:01:43.410413Z","iopub.status.idle":"2024-02-16T21:01:43.492178Z","shell.execute_reply.started":"2024-02-16T21:01:43.410366Z","shell.execute_reply":"2024-02-16T21:01:43.490780Z"},"trusted":true},"execution_count":31,"outputs":[]},{"cell_type":"code","source":"input_shape = train_X_dense.shape","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.494789Z","iopub.execute_input":"2024-02-16T21:01:43.496172Z","iopub.status.idle":"2024-02-16T21:01:43.501444Z","shell.execute_reply.started":"2024-02-16T21:01:43.496107Z","shell.execute_reply":"2024-02-16T21:01:43.499838Z"},"trusted":true},"execution_count":32,"outputs":[]},{"cell_type":"code","source":"print(input_shape)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.502742Z","iopub.execute_input":"2024-02-16T21:01:43.503128Z","iopub.status.idle":"2024-02-16T21:01:43.517962Z","shell.execute_reply.started":"2024-02-16T21:01:43.503085Z","shell.execute_reply":"2024-02-16T21:01:43.517029Z"},"trusted":true},"execution_count":33,"outputs":[{"name":"stdout","text":"(12906, 1060)\n","output_type":"stream"}]},{"cell_type":"code","source":"#from sklearn.decomposition import PCA\n#pca = PCA(.99)\n#pca.fit(X_train)\n#pca.fit(X_test)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.519109Z","iopub.execute_input":"2024-02-16T21:01:43.520061Z","iopub.status.idle":"2024-02-16T21:01:43.531933Z","shell.execute_reply.started":"2024-02-16T21:01:43.520029Z","shell.execute_reply":"2024-02-16T21:01:43.530570Z"},"trusted":true},"execution_count":34,"outputs":[]},{"cell_type":"code","source":"#X_train = pca.transform(X_train)\n#X_test = pca.transform(X_test)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.533718Z","iopub.execute_input":"2024-02-16T21:01:43.534066Z","iopub.status.idle":"2024-02-16T21:01:43.544634Z","shell.execute_reply.started":"2024-02-16T21:01:43.534038Z","shell.execute_reply":"2024-02-16T21:01:43.543317Z"},"trusted":true},"execution_count":35,"outputs":[]},{"cell_type":"code","source":"from keras import layers\n\nfrom kerastuner.tuners import RandomSearch","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:43.546501Z","iopub.execute_input":"2024-02-16T21:01:43.546925Z","iopub.status.idle":"2024-02-16T21:01:59.313644Z","shell.execute_reply.started":"2024-02-16T21:01:43.546873Z","shell.execute_reply":"2024-02-16T21:01:59.312458Z"},"trusted":true},"execution_count":36,"outputs":[{"name":"stderr","text":"2024-02-16 21:01:45.933681: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n2024-02-16 21:01:45.933837: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n2024-02-16 21:01:46.096196: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n/tmp/ipykernel_33/2012623038.py:3: DeprecationWarning: `import kerastuner` is deprecated, please use `import keras_tuner`.\n  from kerastuner.tuners import RandomSearch\n","output_type":"stream"}]},{"cell_type":"code","source":"import keras_tuner\nimport keras","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:59.314952Z","iopub.execute_input":"2024-02-16T21:01:59.315629Z","iopub.status.idle":"2024-02-16T21:01:59.323954Z","shell.execute_reply.started":"2024-02-16T21:01:59.315596Z","shell.execute_reply":"2024-02-16T21:01:59.322693Z"},"trusted":true},"execution_count":37,"outputs":[]},{"cell_type":"markdown","source":"# I removed missing values, standarized and scaled my numeric values and one-hot encoded the categorical features. I'm now ready to tune my model. I tune the number of layers, the dropout rate and activation function ","metadata":{}},{"cell_type":"code","source":"def build_model(hp):\n    model = keras.Sequential()\n    model.add(layers.Dense(units=hp.Int('units_input', min_value=32, max_value=512, step=32), \n                           input_dim=X_train.shape[1], \n                           activation='relu'))  # First layer with input_dim=83\n    for i in range(hp.Int('num_layers', 1, 3)):\n        model.add(layers.Dense(units=hp.Int('units_' + str(i), 32, 512, 32),\n                               activation=hp.Choice('activation_' + str(i), ['relu', 'tanh'])))\n        if hp.Choice('dropout_' + str(i), [0, 1]):\n            model.add(layers.Dropout(rate=hp.Float('dropout_rate_' + str(i), 0.1, 0.5, step=0.1)))\n    model.add(layers.Dense(1, activation='sigmoid'))\n    model.compile(optimizer=hp.Choice('optimizer', ['adam', 'sgd']),\n                  loss='binary_crossentropy',\n                  metrics=[keras.metrics.AUC(name='auc')])\n    return model","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:59.325643Z","iopub.execute_input":"2024-02-16T21:01:59.326107Z","iopub.status.idle":"2024-02-16T21:01:59.341576Z","shell.execute_reply.started":"2024-02-16T21:01:59.326065Z","shell.execute_reply":"2024-02-16T21:01:59.340432Z"},"trusted":true},"execution_count":38,"outputs":[]},{"cell_type":"code","source":"tuner = RandomSearch(\n    build_model,\n    objective=keras_tuner.Objective(\"val_auc\", direction=\"max\"),\n    max_trials=100,\n    executions_per_trial=3,\n    directory='.',\n    project_name='90Diag')","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:59.343240Z","iopub.execute_input":"2024-02-16T21:01:59.344516Z","iopub.status.idle":"2024-02-16T21:01:59.502070Z","shell.execute_reply.started":"2024-02-16T21:01:59.344474Z","shell.execute_reply":"2024-02-16T21:01:59.500787Z"},"trusted":true},"execution_count":39,"outputs":[]},{"cell_type":"code","source":"tuner.search_space_summary()","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:59.503283Z","iopub.execute_input":"2024-02-16T21:01:59.503615Z","iopub.status.idle":"2024-02-16T21:01:59.511118Z","shell.execute_reply.started":"2024-02-16T21:01:59.503588Z","shell.execute_reply":"2024-02-16T21:01:59.509894Z"},"trusted":true},"execution_count":40,"outputs":[{"name":"stdout","text":"Search space summary\nDefault search space size: 6\nunits_input (Int)\n{'default': None, 'conditions': [], 'min_value': 32, 'max_value': 512, 'step': 32, 'sampling': 'linear'}\nnum_layers (Int)\n{'default': None, 'conditions': [], 'min_value': 1, 'max_value': 3, 'step': 1, 'sampling': 'linear'}\nunits_0 (Int)\n{'default': None, 'conditions': [], 'min_value': 32, 'max_value': 512, 'step': 32, 'sampling': 'linear'}\nactivation_0 (Choice)\n{'default': 'relu', 'conditions': [], 'values': ['relu', 'tanh'], 'ordered': False}\ndropout_0 (Choice)\n{'default': 0, 'conditions': [], 'values': [0, 1], 'ordered': True}\noptimizer (Choice)\n{'default': 'adam', 'conditions': [], 'values': ['adam', 'sgd'], 'ordered': False}\n","output_type":"stream"}]},{"cell_type":"code","source":"tuner.search(X_train, y_train,\n          epochs=10,\n          validation_data=(X_test, y_test),\n          batch_size=128)","metadata":{"execution":{"iopub.status.busy":"2024-02-16T21:01:59.513014Z","iopub.execute_input":"2024-02-16T21:01:59.513407Z"},"trusted":true},"execution_count":null,"outputs":[{"name":"stdout","text":"Trial 82 Complete [00h 00m 36s]\nval_auc: 0.7551438808441162\n\nBest val_auc So Far: 0.7793110807736715\nTotal elapsed time: 00h 44m 54s\n\nSearch: Running Trial #83\n\nValue             |Best Value So Far |Hyperparameter\n416               |448               |units_input\n1                 |1                 |num_layers\n384               |416               |units_0\nrelu              |tanh              |activation_0\n1                 |1                 |dropout_0\nsgd               |adam              |optimizer\n352               |128               |units_1\nrelu              |tanh              |activation_1\n1                 |1                 |dropout_1\n0.5               |0.5               |dropout_rate_0\n0.4               |0.1               |dropout_rate_1\n128               |32                |units_2\nrelu              |tanh              |activation_2\n1                 |1                 |dropout_2\n0.3               |0.3               |dropout_rate_2\n\nEpoch 1/10\n81/81 [==============================] - 2s 15ms/step - loss: 0.6765 - auc: 0.5146 - val_loss: 0.6572 - val_auc: 0.5630\nEpoch 2/10\n81/81 [==============================] - 1s 11ms/step - loss: 0.6605 - auc: 0.5476 - val_loss: 0.6511 - val_auc: 0.5988\nEpoch 3/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6525 - auc: 0.5841 - val_loss: 0.6461 - val_auc: 0.6322\nEpoch 4/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6472 - auc: 0.6047 - val_loss: 0.6410 - val_auc: 0.6536\nEpoch 5/10\n81/81 [==============================] - 1s 11ms/step - loss: 0.6402 - auc: 0.6306 - val_loss: 0.6358 - val_auc: 0.6752\nEpoch 6/10\n81/81 [==============================] - 1s 11ms/step - loss: 0.6320 - auc: 0.6576 - val_loss: 0.6302 - val_auc: 0.6905\nEpoch 7/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6290 - auc: 0.6628 - val_loss: 0.6244 - val_auc: 0.7017\nEpoch 8/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6201 - auc: 0.6884 - val_loss: 0.6183 - val_auc: 0.7117\nEpoch 9/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6119 - auc: 0.7048 - val_loss: 0.6114 - val_auc: 0.7215\nEpoch 10/10\n81/81 [==============================] - 1s 10ms/step - loss: 0.6043 - auc: 0.7184 - val_loss: 0.6042 - val_auc: 0.7296\nEpoch 1/10\n39/81 [=============>................] - ETA: 0s - loss: 0.6933 - auc: 0.4775","output_type":"stream"}]},{"cell_type":"code","source":"#batch_size = range(32, 256, 32)\n#epochs = [10, 20, 30]\n#\n#for bs in batch_size:\n#    for en in epochs:\n#        tuner.search(X_train, y_train,\n#             epochs=en,\n#             validation_data=(X_test, y_test),\n#             batch_size=bs)\n#        print(f'Best results for {bs} batch size and {en} number epochs are:\\n')\n#        print(tuner.results_summary(), '\\n')\n        ","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"tuner.results_summary()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Get the best model\nbest_model = tuner.get_best_models(num_models=1)[0]\n\n# Make predictions on the test data\npredictions = best_model.predict(X_test)\n\nprint(predictions)\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.metrics import roc_auc_score","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"score_proba = roc_auc_score(y_test, predictions)\nprint(score_proba)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#test_X_dense = pca.transform(test_X_dense)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sub_proba = best_model.predict(test_X_dense)\ntest['DiagPeriodL90D'] = sub_proba\nsub = test[['patient_id','DiagPeriodL90D']]\nsub.head()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sub.to_csv('submission.csv', index=False)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}