Identification of handwritten digit using svm algorithm comparing random forest to improve accuracy

Abstract

Aim:The main objective of this paper is to recognize the handwritten digits  with the help of Novel Support Vector Machine and Random Forest algorithms. Materials and Methods: The datasets extracted for the SKLEARN module using python language which has around 70000 sample points. Novel Support Vector Machine predicts the output for dependent variable and independent  variable. Sample count for group 1 is 20 and sample count for group group is 20. Total sample size is 40 for both groups using  Gpower as 80%. Results: Novel Support Vector Machine comes up with the mean accuracy when contrasted with the Random Forest algorithm. Ultimately the Novel Support Vector Machine (SVM) pops up with a better value  than the Random Forest (RF) algorithm. The two algorithms SVM and RF are statistically satisfied with the independent sample T-test value (p=.001) with a confidence level of 95%. (8*8) pixel size image will get through Image Processing. After the image is converted into digital format then Image Detection happens to find the digit which is in pixel size format. Conclusion: Within the limits of the study the Novel Support Vector Machine has better significant accuracy value than Random Forest.