Online Updating Algorithms of Statistical Methods for Big Data

In this paper, we discuss online updating algorithms for Big Data. One of the main challenges of Big Data is the limitation of data storage. In the Big Data stream environment, online computation sometimes requires fast updates without the use of historical data. The focus of this research is on efficient online update algorithms for basic statistical computations, including mean, variance, covariance, skewness, kurtosis, confidence interval, test statistic, and linear regression. We demonstrate the implementation of R Language through a linear regress example.