Support Vector Machine is outputting two zeros in Confusion Matrix

Question

I am working on an ML project to predict a particular boolean column. I am using SVM's in SKlearn. I have some features I am trying to use as my descriptive features. These features contain integers (label encoding) and have a max of three unique values. The confusion matrix and classification report that is produced after I make the prediction is down below. I can't understand why it is classifying zero results. Any help? image of code here

The code below is using one descriptive feature that contains three unique integers, and the target feature is a boolean column.

from sklearn.metrics import confusion_matrix
confusion_matrix(y_test, y_pred)
array([[9309,    0],
       [8896,    0]], dtype=int64)

from sklearn.metrics import classification_report
print(classification_report(y_test, y_pred))
      precision    recall  f1-score   support

           0       0.51      1.00      0.68      9309
           1       0.00      0.00      0.00      8896

    accuracy                           0.51     18205

Here is the encoding and the training

encoder = preprocessing.LabelEncoder()
df['Gaze'] = encoder.fit_transform(df['Gaze'])
df['Blink'] = encoder.fit_transform(df['Blink'])
df['Brows'] = encoder.fit_transform(df['Brows'])
df['QuestionType'] = encoder.fit_transform(df['QuestionType'])


y = df[['QuestionType']]
X = df[['Blush']]

from sklearn.svm import SVC
clf = SVC(kernel='linear')

from sklearn.preprocessing import MinMaxScaler
scaling = MinMaxScaler(feature_range=(-1,1)).fit(X_train)
X_train = scaling.transform(X_train)
X_test = scaling.transform(X_test)

clf.fit(X_train, y_train)
y_score = clf.decision_function(X_test)
y_pred = clf.predict(X_test)

1) include your code in the question instead of a link to an image. 2) your classifier is effectively `model = lambda x: 0` — erip, Mar 19 '20 at 19:02
Thank you! It's my first question so still getting used to this :D — LB6221, Mar 19 '20 at 19:07
Welcome. :) Can you also include the code that featurizes your inputs? It seems like your model isn't learning anything, so there's a good chance that's the problem. — erip, Mar 19 '20 at 19:08
Your code isn't complete, but it looks like your X is not encoded. — erip, Mar 19 '20 at 20:25
use LabelEncoder() to encode blush, encode.fit_transform(df['blush']) — Golden Lion, Mar 09 '21 at 00:38

Support Vector Machine is outputting two zeros in Confusion Matrix

0 Answers0