Statistics in Big Data
摘要
Technological advances in biomedicine, computing, and storage have led to an explosion of digital information and present new challenges in data acquisition, processing, management, transferring, and analysis. The value of big data lies in the analytical use of its information to generate knowledge and action. The goal of big data analytics is to extract knowledge from the data to draw conclusions and make decisions. The purpose of this article is to present a view of prospects of statistics in the context of big data analytics. Statistics is a very old discipline for data analysis and data inference using methods based on probability theory. Statistics and data mining techniques that are useful for big data analytics include: significance testing, classification, re-gression/prediction, cluster analysis, association rule learning, anomaly detection, and visualization. Statistical analysis provides a scientific justification to move from data to knowledge to action, and is essential to big data analytics. In addition, big data analytics requires good computer skills in information processing and programming skills as well as knowledge expertise that can be applied to the domain of applications. Statisticians can serve a leadership role in the big data movement.