Paper Title
Variance Matrix: A Solution to Overfitting

Abstract
This paper presents an alternative method to solve overfitting in Machine Learning specifically in Linear Regression. This method is based on adding noise to the system and uses data augmentation to solve overfitting. We constantly fluctuate our training data during train time so that that the Machine Learning Model doesn’t overfit the respective data. The order of variance can be tuned using a hyperparameter. Keywords - Overfitting, Variance Matrix, Machine Learning, Alternative to Regularisation