Data centres have become a major part of computing, and with the advent of cloud computing their demand has soared. The increase in demand for cloud services has led to a data centre boom, however, the data centres tend to consume large amounts of power. The advent of Green computing has led to various researches into how to make computing on a large scale more sustainable. This has led to the evolution of power consumption prediction researches that are meant to help ease the use of power by data centres. In this regard, this research aims to look at ways to cope with the power consumption through adoption of Deep Learning to assist with feature selection. This method aims to look beyond the prior researches into power consumption which only looked at certain factors mainly consumption by the server and not the whole data centre. Key to this whole research area are the following phases: (i) performance monitoring and energy-related feature acquisition; (ii) essential feature selection; and (iii) model establishment and optimization.

Keywords : Green computing, Deep Learning, Feature Selection