Incremental partial least squares analysis of big streaming data

作者：

Highlights：

• We propose a two-stage Incremental PLS (IPLS) dimension reduction method.

• IPLS has low time complexity, linear with both the numbers of samples and features.

• Empirical results show IPLS performs better than some state-of-the-arts methods.

摘要

Highlights•We propose a two-stage Incremental PLS (IPLS) dimension reduction method.•IPLS has low time complexity, linear with both the numbers of samples and features.•Empirical results show IPLS performs better than some state-of-the-arts methods.

论文关键词：Feature extraction,Incremental learning,Large-scale data,Partial least squares,Streaming data

论文评审过程：Received 22 January 2014, Revised 18 May 2014, Accepted 31 May 2014, Available online 9 June 2014.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.05.022