Priprava in uporaba ekspertnega predznanja za avtomatizirano modeliranje vodnih ekosistemov

Abstract

Keywords

Data

Language:	Slovenian
Year of publishing:	2005
Source:	Ljubljana
Typology:	2.08 - Doctoral Dissertation
Organization:	UL FGG - Faculty of Civil and Geodetic Engineering
Publisher:	[N. Atanasova]
UDC:	004.925:574.5(043.3)
COBISS:	2751585
Views:	1918
Downloads:	551
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Preparation and use of the domain expert knowledge for automated modelling of aquatic ecosystems
Secondary abstract:	This thesis is concerned with automated modelling (AM) of aquatic ecosystems. The method used here integrates the two basic principles of modelling, i.e., empirical or data-driven in theoretical or modelling by using the expert background knowledge. The integration of empirical in theoretical modelling is based on the use of the background knowledge in the procedure of model induction from measured data. The theoretical knowledge that guides the process of model induction includes a knowledge library of generalised knowledge from a specific domain in a task specification of the observed system. The thesis is divided into two parts. The first part deals with elaboration of knowledge library in the domain of modelling of aquatic ecosystems. The library includes knowledge about food web modeling by following the mass conservation principle. The knowledge is formalized in terms of (1) taxonomy of variable types, (2) basic processes that govern the behavior of aquatic ecosystems, (3) alternative models of the basic processes, and (4) knowledge how to combine models of individual processes into a model of the entire ecosystem. We evaluated the generality of the knowledge in the library through reconstruction of three wellknown models of different complexity. Thus, we showed that such formalization of the modelling knowledge provides a solid unifying framework for both handcrafting ecological models as well as their automated induction from measured data. In the second part we applied the developed library in the AM method on four real world domains. Using the measurements and the background knowledge we constructed models for each domain. The models were evaluated according to their accuracy and transparency. We tackled the following domains: Lake Glumsoe (Danmark), Lagoon of Venice (Italy), Lake Kasumigaura (Japan), and Lake of Bled (Slovenia). The quality of the models is above all dependant on (1) the knowledge in the library, (2) the quality of the measurements, (3) ecosystem complexity, and (4) the expert knowledge introduced in the induction procedure.
Secondary keywords:	Vodni ekosistemi;Disertacije;Modeliranje;
File type:	application/pdf
Type (COBISS):	Dissertation
Thesis comment:	Univ. Ljubljana, Fak. za gradbeništvo in geodezijo
Pages:	V, 60 str., [113] str. pril.
Type (ePrints):	thesis
Title (ePrints):	Preparation and use of the domain expert knowledge for automated modelling of aquatic ecosystems
Keywords (ePrints):	vodni ekosistem;matematično modeliranje;konceptualno modeliranje;dinamični sistemi;avtomatizirano modeliranje;strojno učenje;domenska knjižnica znanja.
Keywords (ePrints, secondary language):	aquatic ecosistems;mathematical modelling;conceptual modelling;dynamic systems;automated modelling;machine learning;domain knowledge library.
Abstract (ePrints):	Naloga se ukvarja z avtomatiziranim modeliranjem (AM) vodnih ekosistemov. Uporabljena metoda AM (LAGRAMGE) združuje dva osnovna principa modeliranja, t.j. gradnja modelov iz podatkov (empirično), tako kot večina orodij AM in modeliranje z uporabo področnega (teoretičnega) znanja. Združitev teoretičnega in empiričnega pristopa k modeliranju temelji na vpeljavi področnega predznanja v postopek indukcije modelov iz podatkov. Teoretično znanje se upošteva v obliki knjižnice posplošenega znanja iz domene. Vsebinsko je naloga je razdeljena v dva dela. V prvem delu se ukvarjamo z izdelavo posplošene knjižnice znanja za področje modeliranja vodnih ekosistemov. Natančneje, se zajeto znanje nanaša na modeliranje vodnih ekosistemov z upoštevanjem principa masnih bilanc. Posplošeno znanje o dinamiki sistema je formalizirano preko vpeljave (1) generičnih tipov sistemskih spremenljivk, (2) generičnih osnovnih procesov, ki delujejo na spremenljivke, (3) alternativnih modelov osnovnih procesov in (4) znanja o kombiniranju procesov v model celotnega sistema. Ovrednotili smo splošnost znanja v izdelani knjižnici. Z uporabo predznanja v knjižnici smo zapisali več znanih in uveljavljenih modelov vodnih ekosistemov. Tako smo pokazali (poleg splošnosti zajetega znanja), da ustrezno formalizirano znanje omogoča poenoten modularni pristop tako k 'ročni' gradnji modelov kot tudi avtomatski indukciji modelov iz meritev. Drugi del naloge se ukvarja z uporabo metode avtomatiziranega modeliranja (LAGRAMGE), ki zdaj vključuje razvito knjižnico znanja, na realnih podatkih. Z uporabo merjenih podatkov in knjižnice smo zgradili modele, ter jih ovrednotili glede na natančnost in razumljivost oz. transparentnost. Obravnavali smo štiri domene: jezero Glumso, Beneška Laguna, jezero Kasumigaura in Blejsko jezero. Kvaliteta odkritih modelov je odvisna predvsem od (1) znanja zajetega v knjižnici, (2) kvalitete podatkov, (3) kompleksnosti ekosistema in (4) ekspertnega znanja, ki ga vnesemo v postopek odkrivanja modela.
Abstract (ePrints, secondary language):	This thesis is concerned with automated modelling (AM) of aquatic ecosystems. The method used here integrates the two basic principles of modelling, i.e., empirical or data-driven in theoretical or modelling by using the expert background knowledge. The integration of empirical in theoretical modelling is based on the use of the background knowledge in the procedure of model induction from measured data. The theoretical knowledge that guides the process of model induction includes a knowledge library of generalised knowledge from a specific domain in a task specification of the observed system. The thesis is divided into two parts. The first part deals with elaboration of knowledge library in the domain of modelling of aquatic ecosystems. The library includes knowledge about food web modeling by following the mass conservation principle. The knowledge is formalized in terms of (1) taxonomy of variable types, (2) basic processes that govern the behavior of aquatic ecosystems, (3) alternative models of the basic processes, and (4) knowledge how to combine models of individual processes into a model of the entire ecosystem. We evaluated the generality of the knowledge in the library through reconstruction of three wellknown models of different complexity. Thus, we showed that such formalization of the modelling knowledge provides a solid unifying framework for both handcrafting ecological models as well as their automated induction from measured data. In the second part we applied the developed library in the AM method on four real world domains. Using the measurements and the background knowledge we constructed models for each domain. The models were evaluated according to their accuracy and transparency. We tackled the following domains: Lake Glumsoe (Danmark), Lagoon of Venice (Italy), Lake Kasumigaura (Japan), and Lake of Bled (Slovenia). The quality of the models is above all dependant on (1) the knowledge in the library, (2) the quality of the measurements, (3) ecosystem complexity, and (4) the expert knowledge introduced in the induction procedure.
Keywords (ePrints, secondary language):	aquatic ecosistems;mathematical modelling;conceptual modelling;dynamic systems;automated modelling;machine learning;domain knowledge library.
ID:	8312785