EURASIP Journal on Wireless Communications and Networking

Table 3 Workload prediction result of baseline methods and our approach on Alibaba cluster-trace-2018

From: A novel approach to workload prediction using attention-based LSTM encoder-decoder network in cloud environment

Model	Alibaba×10⁻²
Model	MAE	RMSE	MAPE	RMSSE
ARIMA	7.016	8.337	33.126%	8.272
PSR + EA-GMDH	7.021	8.346	33.174%	8.311
LSTM	5.756	6.903	28.031%	6.014
GRUED	4.371	5.211	24.362%	4.432
Our model	3.520	4.134	19.529%	3.815

Back to article page