Author(s)

Yonghong Zhang

Corresponding Author

Yonghong Zhang

Abstract

We have entered the era of big data in modern and contemporary times, and after entering the era of big data, more and more fields have to deal with massive and high-speed data in real time. How to extract these large data streams and effectively transform them into practical information is very important for all walks of life. The traditional batch machine learning technology, which we are familiar with, has many restrictive factors in the application and analysis of large data in the contemporary era, while the current online learning technology adopts the streaming computing mode. Its main advantage is that the real-time analysis and calculation of data can be carried out directly in memory, so as to achieve the streaming mode. Data learning provides advantageous methods and tools, and it also introduces the background and motivation of big data analysis. It shows the latest and classical online learning algorithms and methods. Therefore, this latest online learning system is very likely and hopeful to solve all kinds of big data mining tasks. Challenges, the main technology and content include these aspects: first, linear model online learning; second, non-traditional online learning methods. All these methods give more detailed models or pseudo-codes as far as possible, so as to discuss a large number of machine learning for large data analysis. Key issues in use.

Keywords

Stream data, big data analysis, online learning