Efficiency Analysis of Predicting Handwritten Equation Solver Using CNN Over SVM Algorithm

In addition to learning about the effectiveness of revolutionary Convolutional Neural Network techniques and SVMs, the main objective of this research is to provide a novel method for Handwritten Equation Recognition. The sample sizes used to estimate the accuracy percentages are <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{N}=\mathbf{1 0}$</tex> for the CNN technique and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{N}=\mathbf{1 0}$</tex> for SVM, with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{2 0}$</tex> iterations. Using clincalc.com, the sample size was calculated using a threshold value of 0.05 percent and a confidence interval of 95 percent, a G power of 80 percent, a minimum power of 0.8, and a maximum tolerable error of 0.5. The novel CNN algorithm outperforms the SVMs technique in terms of accuracy (93.65 %). (92.85 %). Keeping the <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$G$</tex> power at 80 %, a 95 % confidence interval and a threshold of 0.05. The novel approach to handwritten equation recognition differs from the conventional method by 0.145 in terms of statistical significance. The results of the study are statistically insignificant, according to the value of the independent sample T-test. The outcomes showed that in handwritten equation recognition, the novel CNN approach performs better than SVMs.