Wyniki wyszukiwania - BazTech

1

Controllling simulation experiment design for agent-based models using tree representation of parameter space

Szufel P, Kamiński B., Wojewnik P.

Foundations of Computing and Decision Sciences

|

2013

|

Vol. 38, No. 4

277--298

EN

An important aspect of the simulation modelling process is sensitivity analysis. In this process, agent-based simulations often require analysis of structurally different parameter specifications – the parameters can be represented as objects and the object-oriented simulation configuration leads to nesting of simulation parameters. The nested parameters are naturally represented as a tree rather than a flat structure. The standard tools supporting multi-agent simulations only allow only the representation of the parameter space as a Cartesian product of possible parameter values. Consequently, their application for the required tree representation is limited. In this paper an approach to tree parameter space representation is introduced with an XML-based language. Furthermore, we propose a set of tools that allows one to manage parameterization of the simulation experiment independently of the simulation model.

2

ICA method and RBF neural network for prediction improvement

Szupiluk R., Siwek K., Wojewnik P., Ząbkowski T.

Przegląd Elektrotechniczny

|

2007

|

R. 83, nr 11

57-59

EN

In this paper we present a novel method for integration the prediction results by finding common latent components via independent component analysis. The latent components can have constructive or destructive influence on particular prediction results. After the elimination of the deconstructive signals we rebuilt the improved predictions using RBF neural networks. We check the method validity on the electricity load prediction task.

PL

W artykule przedstawiono nową metodę pozwalającą na łączenie wyników predykcji poprzez poszukiwanie ukrytych wspólnych składowych przy zastosowaniu procedury analizy składowych niezależnych. Składowe ukryte mogą mieć pozytywny lub negatywny wpływ na wyniki predykcji. Po wyeliminowaniu składowych niekorzystnych poprawiono model wykorzystując sztuczną sieć neuronową z radialną funkcją bazową. Poprawność metody sprawdzono na przykładzie predykcji poboru mocy.

3

Independent Component Analysis for Modelling Improvement

Szupiluk R., Wojewnik P., Ząbkowski T.

Schedae Informaticae

|

2004

|

Vol. 13

151-164

EN

A common problem encountered in such disciples as statistics, data analysis, signal processing, data mining and time series analysis is finding a suitable model to explore existing dependencies. There are many models with different advantages and it happens that different good criteria indicate different models as an optimal solution. We need good criteria to choose the best model. But how to choose the best criterion? To avoid this disadvantage we propose an Independent Component Analysis for adopting many solutions from a large set of competitive models. Such a transformation seems to capture the essential structure of data solutions in many applications.