Handwritten digit recognition

  • The purpose of this project is to compare classify images to recognize the handwritten digits.
  • This task was performed using different classifiers, namely, logistic regression classifier, Neural Networks, Support Vector Machine and Random Forests.
  • The results of these classifiers were compared and analyzed by generating a confusion matrix.
  • The predictions of all these classifiers were combined to design a combined classifier that works on a majority voting system.

Find the entire project here: