2026-09-21T19:52:48Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/3370 2022-03-24T23:11:34Z 1582963302567:1597824273898 user-ucsy

Cloud Data Center Resource Demand Prediction Model Development on Apache Spark Than, Moh Moh Thein, Thandar Dynamic resource allocation in cloud datacenters is a challenging problem. Resourceprediction is a key feature for on-demand resourceplanning and efficient resource management ofdynamic workload. This requires a highly accuratedemand prediction. Hyper-parameter optimizationcan largely affect the performance of the predictionmodel. The process of identifying the optimalparameters for a machine learning (ML) algorithminvolves the search for a broad range of valuecombinations of parameter sets. This paper presentsa resource demand prediction model with the cloudcomputational frameworks Apache™ Hadoop® andApache Spark™. The model is developed on thepowerful ML technique, Decision Tree (DT)algorithm, and hyper-parameter optimization for DTalgorithm is performed to achieve the predictionmodel with high accuracy. The evaluation ofprediction model is conducted on real data centerworkload traces and the evaluation results show thathyper-parameter optimization can save theprediction error significantly. 2019-02-27 http://hdl.handle.net/20.500.12678/0000003370 https://meral.edu.mm/records/3370