Ensembles of Fuzzy Linear Model Trees for the Identification of Multioutput Systems

Aleksovski, Darko; Kocijan, Juš; Džeroski, Sašo

doi:10.1109/tfuzz.2015.2489234

Cited by 11 publications

(3 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In black-box modeling, we conjecture that the appropriate model structure belongs to a general class of structures, such as neural networks and fuzzy models [16] or regression trees [1]. While nonlinear black-box models can achieve highly accurate reconstruction of the observed system behavior, they do not reveal the structure of the observed system.…”

Section: Introductionmentioning

confidence: 99%

Equation Discovery for Nonlinear System Identification

et al. 2020

Self Cite

View full text Add to dashboard Cite

Equation discovery methods enable modelers to combine domain-specific knowledge and system identification to construct models most suitable for a selected modeling task. The method described and evaluated in this paper can be used as a nonlinear system identification method for gray-box modeling. It consists of two interlaced parts of modeling that are computer-aided. The first performs computeraided identification of a model structure composed of elements selected from user-specified domain-specific modeling knowledge, while the second part performs parameter estimation.In this paper, recent developments of the equation discovery method called process-based modeling, suited for nonlinear system identification, are elaborated and illustrated on two continuous-time case studies. The first case study illustrates the use of the process-based modeling on synthetic data while the second case-study evaluates on measured data for a standard system-identification benchmark. The experimental results clearly demonstrate the ability of process-based modeling to reconstruct both model structure and parameters from measured data. Index Terms-Software for system identification, Gray-box modeling, Nonlinear system identification, Equation discovery, Process-based modeling, Knowledge-based identification, Machine learningAll authors are affiliated with the Jožef Stefan Institute,

show abstract

Section: Introductionmentioning

confidence: 99%

Equation Discovery for Nonlinear System Identification

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…As tabelas 31 e 32 comparam os resultados obtidos pelas melhores soluções dos submodelos (sem considerar os submodelos AEIQ-R/SRG/ESN e CMA-ES/CMA-ES/ESN, que geraram algumas soluções instáveis) com os resultados dos métodos de Aleksovski et al (2014Aleksovski et al ( , 2016, que utilizaram as mesmas amostras de teste, onde as métricas RRSE e RRMSE são dadas por:…”

Section: Método Métrica Fitunclassified

“…Considerando o desempenho médio nos 10 experimentos em termos de RRSE, pela tabela 30, pode-se ver que ambas as abordagens que passaram pela otimização do reservatório conseguiram gerar resultados superiores aos resultados médios dos métodos apresentados porAleksovski et al (2016), ainda que os submodelos AEIQ-CA valores médios de RMSE durante o teste, o que foi devido a soluções instáveis geradas por estes métodos. As figuras 101 e 102 mostram as curvas de evolução dos algoritmos AEIQ-R e CMA-ES nas etapas 1 e 2, exibindo-se a média da aptidão nos 10 experimentos do melhor indivíduo da população, enquanto a tabela 33 exibe os valores de aptidão na obter desempenho superior ao CMA-ES durante todo o processo evolutivo, convergindo para valores menores de erro de validação, porém como visto pela tabela 32, as soluções obtidas pelo submodelo CMA-ES/CMA-ES/ESN foram de melhor qualidade, obtendo melhor desempenho durante o teste.…”

unclassified