Handwritten Digit Recognition Using Machine Learning

Year 2021, Volume: 25 Issue: 1, 65 - 71, 01.02.2021

Rabia Karakaya , Serap Kazan

https://doi.org/10.16984/saufenbilder.801684

Cited By: 2

Abstract

Technology is getting more and more involved in our lives, and so are algorithms. These algorithms speed up work and reduce workload. Especially machine learning algorithms are improving day by day by imitating human behaviours. Handwriting recognition systems are also stand out on this field. In this study, handwriting digit recognition process has been done with algorithms having different working methods. These algorithms are Support Vector Machine (SVM), Decision Tree, Random Forest, Artificial Neural Networks (ANN), K-Nearest Neighbor (KNN) and K- Means Algorithm. The working logic of the handwriting digit recognition process was examined, and the efficiency of different algorithms on the same database was measured. A report was presented by making comparisons on the accuracy.

Keywords

handwritten digit recognition, machine learning, artificial intelligence

References

• [1] Sekerci, M., “Birlesik ve egik Turkce el yazisi tanima sistemi,” Trakya University, Master’s Thesis, 2007.
• [2] Yilmaz, B., “Ogrenme guclugu ceken cocuklar icin el yazisi tanima ile ogrenmeyi kolaylastirici bir mobil ogrenme uygulamasi tasarimi,” Maltepe University, Master’s Thesis, 2014.
• [3] Dagdeviren, E., “El yazisi rakam tanima icin destek vektor makinelerinin ve yapay sinir aglarinin karsilastirmasi,” Istanbul University, Master’s Thesis, 2013.
• [4] Assegie, T. A., Nair, P. S., “Handwritten digits recognition with decision tree classification: a machine learning approach,” International Journal of Electrical and Computer Engineering (IJECE), Indonesia, 446-4451, 2019.
• [5] Arica, N., Vural, Y., F.T., “An overview of character recognition focused on offline handwriting,” IEE Transactions on Systems Man and Cybernetics, 2001.
• [6] MacKenzie, S., Tanaka-Ishii, K., “Text entry systems,” Morgan Kaufmann Publishing, 123-137, 2007.
• [7] Gunn, S.R., “Support vector machines for classification and regression,” 1998.
• [8] Safavian, S.R., Landgrebe, D., “A survey of decision tree classifier methodology,” IEE Transactions on Systems Man and Cybernetics, 1991.
• [9] Géron, A., “Hands-On machine learning with Scikit-Learn and TensorFlow concepts,” O’Reilly Publishing, 2017.
• [10] Arya, S., Mount, D.M., Netanyahu, N.S., Silverman, R. Wu, A.Y., “An optimal algorithm for approximate nearest neighbor searching in fixed dimension,” Journal of the ACM, 2017.
• [11] Kirmizigul Caliskan, S., Sogukpinar, I., KxKNN : “K-ortalama ve K-en yakin komsu yontemleri ile aglarda nufuz tespiti,” Electronic Journal of Vocational Colleges (Ejovoc) , Kırklareli , 2015.
• [12] Gunes, A., Yigit, T., “Hizlandirilmis destek vektor makineleri ile el yazisi rakamlarin taninmasi,” 20th Signal Processing and Communications Applications Conference, Mugla, 2012.
• [13] Liu, C., Fujisawa, H., “Handwritten digit recognition: Benchmarking of state-of-the-art techniques,” Pattern Recognition, 2003.
• [14] Fialoke, S., “Predicting Digits from their handwritten images, (16.10.2020), http://suruchifialoke.com/2017-06-15-predicting-digits_tensorflow.