This paper presents a novel fuzzy firefly-based intelligent algorithm for load balancing in mobile cloud computing while reducing makespan.The proposed technique implicitly acts intelligently by using inherent traits ...This paper presents a novel fuzzy firefly-based intelligent algorithm for load balancing in mobile cloud computing while reducing makespan.The proposed technique implicitly acts intelligently by using inherent traits of fuzzy and firefly.It automatically adjusts its behavior or converges depending on the information gathered during the search process and objective function.It works for 3-tier architecture,including cloudlet and public cloud.As cloudlets have limited resources,fuzzy logic is used for cloudlet selection using capacity and waiting time as input.Fuzzy provides human-like decisions without using any mathematical model.Firefly is a powerful meta-heuristic optimization technique to balance diversification and solution speed.It balances the load on cloud and cloudlet while minimizing makespan and execution time.However,it may trap in local optimum;levy flight can handle it.Hybridization of fuzzy fireflywith levy flight is a novel technique that provides reduced makespan,execution time,and Degree of imbalance while balancing the load.Simulation has been carried out on the Cloud Analyst platform with National Aeronautics and Space Administration(NASA)and Clarknet datasets.Results show that the proposed algorithm outperforms Ant Colony Optimization Queue Decision Maker(ACOQDM),Distributed Scheduling Optimization Algorithm(DSOA),andUtility-based Firefly Algorithm(UFA)when compared in terms of makespan,Degree of imbalance,and Figure of Merit.展开更多
In this paper,we propose a novel flexible optimization pipeline for determining the optimal adsorption sites,named AUGUR(Aware of Uncertainty Graph Unit Regression).Our model combines graph neural networks and Gaussia...In this paper,we propose a novel flexible optimization pipeline for determining the optimal adsorption sites,named AUGUR(Aware of Uncertainty Graph Unit Regression).Our model combines graph neural networks and Gaussian processes to create a flexible,efficient,symmetry-aware,translation,and rotation-invariant predictor with inbuilt uncertainty quantification.This predictor is then used as a surrogate for a data-efficient Bayesian Optimization scheme to determine the optimal adsorption positions.This pipeline determines the optimal position of large and complicated clusters with far fewer iterations than current state-of-the-art approaches.Further,it does not rely on hand-crafted features and can be seamlessly employed on any molecule without any alterations.Additionally,the pooling properties of graphs allow for the processing of molecules of different sizes by the same model.This allows the energy prediction ofcomputationally demanding systemsby a model trained on comparatively smaller and less expensive ones.展开更多
基金funded by University Grant Commission with UGC-Ref.No.:3364/(NET-JUNE 2015).
文摘This paper presents a novel fuzzy firefly-based intelligent algorithm for load balancing in mobile cloud computing while reducing makespan.The proposed technique implicitly acts intelligently by using inherent traits of fuzzy and firefly.It automatically adjusts its behavior or converges depending on the information gathered during the search process and objective function.It works for 3-tier architecture,including cloudlet and public cloud.As cloudlets have limited resources,fuzzy logic is used for cloudlet selection using capacity and waiting time as input.Fuzzy provides human-like decisions without using any mathematical model.Firefly is a powerful meta-heuristic optimization technique to balance diversification and solution speed.It balances the load on cloud and cloudlet while minimizing makespan and execution time.However,it may trap in local optimum;levy flight can handle it.Hybridization of fuzzy fireflywith levy flight is a novel technique that provides reduced makespan,execution time,and Degree of imbalance while balancing the load.Simulation has been carried out on the Cloud Analyst platform with National Aeronautics and Space Administration(NASA)and Clarknet datasets.Results show that the proposed algorithm outperforms Ant Colony Optimization Queue Decision Maker(ACOQDM),Distributed Scheduling Optimization Algorithm(DSOA),andUtility-based Firefly Algorithm(UFA)when compared in terms of makespan,Degree of imbalance,and Figure of Merit.
基金funding from the project ProperPhotoMile,supported under the umbrella of SOLAR-ERA.NET Cofund 2 by The Spanish Ministry of Science and Education and the AEI under the project PCI2020-112185 and CDTI project number IDI-20210171the Federal Ministry for Economic Affairs and Energy on the basis of a decision by the German Bundestag project number FKZ 03EE1070B and FKZ 03EE1070A+2 种基金the Israel Ministry of Energy with project number 220-11-031.SOLAR-ERA.NET is supported by the European Commission within the EU Framework Program for Research and Innovation HORIZON 2020(Cofund ERA-NET Action,786483)Further,A.G.acknowledges financial support from TUM Innovation Network for Artificial Intelligence powered Multifunctional Material Design(ARTEMIS)funding in the framework of Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)under Germany’s Excellence Strategy-EXC 2089/1-390776260(e-conversion).Lastly,we wish to express our gratitude to Dr.Inigo Iribarren for creating the flowcharts for this work.
文摘In this paper,we propose a novel flexible optimization pipeline for determining the optimal adsorption sites,named AUGUR(Aware of Uncertainty Graph Unit Regression).Our model combines graph neural networks and Gaussian processes to create a flexible,efficient,symmetry-aware,translation,and rotation-invariant predictor with inbuilt uncertainty quantification.This predictor is then used as a surrogate for a data-efficient Bayesian Optimization scheme to determine the optimal adsorption positions.This pipeline determines the optimal position of large and complicated clusters with far fewer iterations than current state-of-the-art approaches.Further,it does not rely on hand-crafted features and can be seamlessly employed on any molecule without any alterations.Additionally,the pooling properties of graphs allow for the processing of molecules of different sizes by the same model.This allows the energy prediction ofcomputationally demanding systemsby a model trained on comparatively smaller and less expensive ones.
