Classification methods for handwritten digit recognition: a survey

HIGHLIGHTS

• who: Ira M. Tubaa et al. from the Singidunum University, Faculty of Informatics and Computing have published the article: Classification methods for handwritten digit recognition: A survey, in the Journal: (JOURNAL)
• what: The paper analyzes, synthesizes and compares the development of different classifiers applied to the handwritten digit recognition problem, from linear classifiers to convolutional neural_networks.
• how: The conclusions about the quality of different classifiers and features are presented. To enable a comparison between different methods for handwritten digit recognition the MNIST dataset was proposed. In (Patel and amp Kalyani 2016) an inverse . . .
  

   

  

  If you want to have access to all the content you need to log in!

  Thanks :)

  Username or Email

  Password

  Remember me

  Lost your password?

  If you don't have an account, you can create one here.