The learning curves of optimisation algorithms, plotting the evolution of the objective vs. runtime spent. can be viewed as a sample of longitudinal data. In this paper we describe mixed-effects modeling, a standard technique in longitudinal data analysis, and give an example of its application to algorithm performance modeling.

Workload characterization is an important and basic step for performance modeling and capacity planning. Traditional approaches for workload characterization often require sophisticated and time-consuming user assistance. In this paper we describe a number of algorithms for automatically classifying workloads and for determining what resources a workload used even when the operating system does...

The Vertical Axis Wind Turbines Darrieus is facing a rapid installations growth due to the interest for a decentralizing generation. The Darrieus aerodynamic behavior analysis is a hard computational task. This paper starts by offering an insight into the Darrieus wind turbines performance prediction and proposes an algorithm based on the Multiple Streamtube modeling as a prediction approach in...

Regular forms of Amdahl’s law and the universal scalability model are functions of the number of processors p alone, therefore modelling only data sets with a single machine varying in its processor count. This paper demonstrates how these performance models can be expressed in terms of intrinsic machine characteristics such as: processor clock; external resource access speed; total threads; et...

A typical characteristic of refrigeration compressor performance tests is their long duration. A reduction in the time periods related to this activity can be achieved using unsteady-state data analysis. This paper presents an original approach to predicting compressor performance using Bayesian networks and a hybrid Fuzzy–Bayesian network. All analysis was performed using real test data. 2011 ...

This paper deals with the heat exchange performance prediction of a counter flow double-tube condenser for binary natural refrigerant mixtures (propane/n-butane & propane/i-butane ). The local characteristics of the heat transfer and pressure drop are calculated using a prediction method developed by authors, and the total pressure drop, the overall heat transfer coefficient and the average hea...

We describe the integration of the EMPA based software tool TwoTowers and the GSPN based software tool GreatSPN, in order to fully realize a multiparadigm approach to the functional and performance modeling and analysis of concurrent and distributed systems. The integration is achieved via an improved semantics mapping EMPA terms onto GSPN models.

Distributed applications running on clusters may be composed of several components with very different performance requirements. The FlowVR middleware allows the developer to deploy such applications and to define communication and synchronization schemes between components without modifying the code. While it eases the creation of mappings, FlowVR does not come with a performance model. Conseq...

This paper describes our experiment and analysis of utilizing prerequisite skill features to improve the predicting of student retention performance. There are two aspects that make this paper interesting. First, instead of focusing on short-team performance, we investigated the student retention performance after a delay of 7 days. We explored several prerequisite skill features that can be ca...

Parallel programs, generated by the supercompiler, VFCS (Vienna Fortran Compilation System), are an application area for the tool, PEPSY (PErformance Prediction System), which we have developed recently. PEPSY derives a performance model from an internal representation of the parallel program in the compiler, automatically, and makes performance analysis by discrete-event simulation. Several mo...