Introduction

Machine learning algorithms are playing their roles in a wide range of industries. For example, classification and regression algorithms are widely used in carriers' high-value customer identification, financial risk evaluation, and traffic prediction scenarios; feature engineering algorithms are used by carriers to extract user credit characteristics and by financial organizations to analyze model features.

Kunpeng BoostKit for Big Data provides a machine learning algorithm library, which has deep optimization in algorithm principles and Kunpeng affinity based on the native Spark algorithm, and therefore has 20 times higher algorithm execution efficiency. The Kunpeng BoostKit for Big Data algorithm library is easy to use. It has the same class and interface definitions as the native Spark algorithm. If customers develop their own applications based on the native Spark algorithm and want to use the BoostKit algorithm library, they do not need to modify any code of the applications. They can directly replace the algorithm library to obtain the performance benefits.

The Kunpeng BoostKit for Big Data algorithm library optimizes multiple algorithms, including classification and regression, recommendation, feature engineering, clustering, and pattern mining, to fit into major application scenarios.