Add better table

Author: cedced19

get_max_scores.py

@@ -14,23 +14,51 @@ def getMaxScoreMethod(label, arr):
             max_score_method = experiment
     return max_score_method
 
+def removeRandomStateParam(model):
+    if (model != None):
+        d=model.get_params()
+        d['random_state'] = None
+        model.set_params(**d)
+    return model
+
+
 def renderMaxScoreMethod(label, arr):
     d = getMaxScoreMethod(label, arr)
     html = "<h3>Method for: maximum of " + label + "</h3>"
     html += "<p><ul>"
     for key, value in d.items():
         if (key == 'oversampler' or key == 'classifier'):
             value = removeRandomStateParam(value)
-        html += "<li>" + key + " "+str(value)+"</li>"
+        if (key == 'cm'): 
+            html += "<li>" + str(key) + ":<br>"  + showCM(value) +"</li>"
+        else:
+            html += "<li>" + key + " " + str(value) + "</li>"
     html+= "</ul></p>"
     return html
 
-def removeRandomStateParam(model):
-    if (model != None):
-        d=model.get_params()
-        d['random_state'] = None
-        model.set_params(**d)
-    return model
+def showCM(cm):
+    cm /= cm.sum()
+    cm = cm*100
+    cm = cm.round(2)
+    html = "<table style=''>\
+    <tr>\
+        <th></th>\
+        <th style='font-weight: bold;'>Negative predicted</th>\
+        <th style='font-weight: bold;'>True predicted</th>\
+    </tr>\
+    <tr>\
+        <td style='font-weight: bold;'>Negative known</td>\
+        <td>TN: {}%</td>\
+        <td>FP: {}%</td>\
+    </tr>\
+    <tr>\
+        <td style='font-weight: bold;'>True known</td>\
+        <td>FN: {}%</td>\
+        <td>TP: {}%</td>\
+    </tr>\
+    </table>".format(cm[0,0], cm[1,0], cm[0,1], cm[1,1])
+    return html
+
 
 def renderHTML(name,arr,dataset):
     if (len(arr) == 0): 

main.py

@@ -18,6 +18,8 @@
 from imblearn.over_sampling import RandomOverSampler, SMOTE, BorderlineSMOTE, SMOTEN
 from imblearn.under_sampling import RandomUnderSampler
 from sklearn.ensemble import RandomForestClassifier
+from sklearn.svm import LinearSVC
+from sklearn.svm import SVC
 
 input_path = "datasets/"
 output_path = "results/"
@@ -188,13 +190,33 @@ def main():
             (
                 'GBM',GradientBoostingClassifier(),
                 [{
-                    'n_estimators': [50, 100, 200]
+                    'n_estimators': [50, 100, 200],
+                    'loss': ['deviance', 'exponential'],
+                    'max_features': ['sqrt', 'log2']
                 }]
             ),(
                 'KNN',KNeighborsClassifier(),
                 [{
                     'n_neighbors': [3,5,8]
                 }]
+            ),(
+                'SVC',SVC(),
+                [{
+                    'kernel': ['linear', 'poly', 'sigmoid']
+                }]
+            ),(
+                'LinearSVC-l1',LinearSVC(),
+                [{
+                    'penalty': ['l1'],
+                    'loss': ['squared_hinge'],
+                    'dual': [False]
+                }]
+            ),(
+                'LinearSVC-l2',LinearSVC(),
+                [{
+                    'penalty': ['l2'],
+                    'loss': ['hinge', 'squared_hinge']
+                }]
             ),(
                 'RandomForestClassifier', RandomForestClassifier(),
                 [{
@@ -227,12 +249,6 @@ def main():
                 [{
                     'k_neighbors': [3,5,20]
                 }]
-            ),
-            (
-                'B2-SMOTE', BorderlineSMOTE(kind='borderline-2'),
-                [{
-                    'k_neighbors': [3,5,20]
-                }]
             )
         ]