Offline handwritten signature verification system using random forest classifier

This research was conducted to find a feasible solution to verify hand written signatures. The scope has been narrowed down to offline signatures which contains static inputs and outputs. Several classification methods such as Multinomial Naive Bayes Classifier (MNBC), Bernoulli Naive Bayes Classifier (BNBC), Logistic Regression Classifier (LRC), Stochastic Gradient Descent Classifier (SGDC) and Random Forest Classifier (RFC) were implemented to identify the most suitable classifier to verify hand written signatures. The classifiers were trained and tested using a signature database available for the public use. The best performance was obtained from RFC with and accuracy score 0.6. For an average, the system created has been successful in verifying signature images provided with a considerable accuracy level.