Python code for Spam Ham classification of e-mails using Naive-Bayes Classifier and Logistic Regression methods, with and without stop words.
Run the following command -
python naiveBayes.py 'spam training folder' 'ham training folder' 'spam test folder' 'ham test folder' 'stop Word file location'
Run the following command -
python logisticRegressionMain.py 'spam training folder' 'ham training folder' 'spam test folder' 'ham test folder' 'learningConst' 'penaly Lambda' 'num of Iterations' 'stopWords location file'
Accuracy without removing stopwords=94.56%
Accuracy after removing stopwords=93.93%
| Learning Const | Lambda | Iterations | Accuracy W/O StopWords | Accuracyafter Removing StopWords |
|---|---|---|---|---|
| 0.01 | 2 | 10 | 92.46 | 89.74 |
| 0.01 | 3 | 20 | 89.12 | 86.82 |
| 0.001 | 1 | 20 | 88.91 | 87.029 |
| 0.0001 | 1 | 30 | 86.40 | 79.07 |
| 0.0001 | 2 | 50 | 87.23 | 81.38 |
| 0.01 | 2 | 10 | 92.05 | 92.25 |