Rishi Khanna
Independent Researcher
India
Abstract
Support Vector Machines (SVMs) have emerged as a powerful supervised learning technique for pattern recognition tasks, including handwritten digit recognition. This manuscript reviews the application of SVMs to the recognition of handwritten digits, focusing exclusively on methods, case studies, and technologies available up to 2017. We begin with an overview of SVM theory and its advantages for high‐dimensional classification. We then examine prominent case studies, notably the MNIST and USPS datasets, illustrating preprocessing and feature extraction techniques such as Histogram of Oriented Gradients (HOG) and Principal Component Analysis (PCA). Next, we identify research gaps in model scalability, feature robustness, and multiclass extension prior to 2017. The methodology section details data preparation, parameter tuning, kernel selection, and validation strategies. Results from various studies are synthesized, demonstrating classification accuracies typically ranging from 98% to 99.5%, depending on feature sets and kernel configurations. Finally, we conclude with insights into best practices and suggestions for future research within the pre‐2017 context.
Keywords
Support Vector Machines, Handwritten Digit Recognition, MNIST, HOG, Kernel Methods
References
- Cortes, C., & Vapnik, V. (1995). Support‐vector networks. Machine Learning, 20(3), 273–297.
- LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient‐based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278–2324.
- Osuna, E., Freund, R., & Girosi, F. (2002). Training support vector machines: an application to face detection. In Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition (pp. 130–136).
- Dalal, N., & Triggs, B. (2005). Histograms of oriented gradients for human detection. In 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) (Vol. 1, pp. 886–893).
- Hsu, C.‐W., & Lin, C.‐J. (2002). A comparison of methods for multiclass support vector machines. IEEE Transactions on Neural Networks, 13(2), 415–425.
- Slaney, M., & Casey, M. (2000). Locality in speech recognition and music classification. In Advances in Neural Information Processing Systems (pp. 506–512).
- Suen, C. Y., & Gader, P. (2003). Nonlinear classification of handwritten digit images using a multi‐layer perceptron. Pattern Recognition, 36(8), 1913–1923.
- Jin, W., & Kumar, M. P. (2001). Improving handwritten digit recognition using support vector machines and principal component analysis. Pattern Recognition Letters, 22(13), 1479–1485.
- Lee, C.‐B., Park, S.‐H., & Kim, D. (2006). Offline handwritten devnagari numeral recognition using support vector machine. In 2006 International Conference on Cognition and Recognition (ICCR) (pp. 225–230).
- Westphal, C., & Bleicher, A. (2004). Directed‐acyclic‐graph SVMs for pattern classification. In 2004 IEEE International Conference on Neural Networks (pp. 1069–1074).
