西华学院K.U.Leuven
Department of Electrical Engineering(ESAT)SISTA
Technical report97-70
DAISY:A databa for identification
of systems∗
B.De Moor,P.De Germ,B.De Schutter,and W.Favoreel
If you want to cite this report,plea u the following reference instead:
B.De Moor,P.De Germ,B.De Schutter,and W.Favoreel,“DAISY:A非主流符号>五年级古诗大全
databa for identification of systems,”Journal A,vol.38,no.3,pp.4–5,
Sept.1997.
ESAT-SISTA
K.U.Leuven
Leuven,Belgium
phone:+32-16-32.17.09(cretary)
fax:+32-16-32.19.70
URL:www.esat.kuleuven.ac.be/sista-cosic-docarch
DAISY:A Databa for Identification of Systems∗
Bart De Moor†Peter De Germ‡Bart De Schutter§Wouter Favoreel¶ESAT/SISTA,Kardinaal Mercierlaan94,B-3001Leuven,Belgium,
tel.:(+32)-(0)16-32.17.09,fax:(+32)-(0)16-32.19.70
{bart.demoor,peter.degerm,bart.deschutter,wouter.favoreel}@esat.kuleuven.ac.be web:www.esat.kuleuven.ac.be/sista
Abstract
We point out the existence of a disturbing deficiency in thefield of system identifica-
tion,namely the fact that many results,published in papers,are not reproducible.In
many cas,datats and time ries,that are ud to illustrate identification methods
and algorithms in the publications,are not freely available.We propo to remedy this
rious deficiency by tting up a publically accessible website,called DAISY,to which
authors can submit datats that are ud to illustrate certain claims and algorithms in
their papers.Several additional benefits are discusd as well.
Keywords:System identification,signal processing,time ries analysis,data analysis,什么炒饭好吃
modeling,datats.
1To measure is
Reproducibility is one of the most basic characteristics of scientific rearch.Yet,in the
fields of data analysis,system identification and signal processing,this very aspect is often neglected or even completely ignored.By this we mean the following:Too often,papers under
review or papers published in conference proceedings and journals,contain an illustration of
a certain algorithm,applied to a given data t.A typical statement is then that’such and
such method works well on such and such datat’.The problem is that this specific datat is
almost always unavailable and inaccessible.The critical reader is confronted with the paradox
that the theoretical derivation of the algorithm in the paper ems to be right but that the
verification of its behavior,when applied to the real datat,is merely impossible.Therefore,
the value of such an illustration of a method when applied to a real datat,is scientifically ∗Work supported by the Flemish Government(BOF(GOA-MIPS),AWI(Bil.Int.Coll.),FWO(projects,
(ICCoS)),IWT(IWT-VCST(CVT),ITA(ISIS),EUREKA(Sinopsys))),the Belgian
Federal Government(IUAP IV-02,IUAP IMechS),the European Commission(HCM(Simonet),TMR(Ala-
pedes),ACTS(Aspect),SCIENCE(ERNSI)),NATO(CRG)and industry(Electrabel).For more details on
the projects,e web-coordinates.dl是谁
†Rearch Associate with the FWO(Fund for Scientific Rearch-Flanders),Associate Professor
氨基比林又叫什么K.U.Leuven;
‡Rearch Assistant with the FWO;
§Senior Rearch Assistant with the FWO;美容英语
¶Rearch Assistant supported by the IWT.
void and may be only aesthetic.However,in many cas,the author of the paper had to go through a lot of trouble to obtain the given datat.Everyone who has been active in experimental work,knows how difficult and time-consuming it is to t up an experiment, obtain measurements,decide onfilters,sampling frequencies,nsors,data-acquisition and When all of this is done,there remains the confrontation of reality with the theoretical framework,which is always bad
on assumptions and hypothes,that never em to be satisfied in practice.The central challenge in system identification and signal processing is precily this confrontation between experimentally obtained measurements and mathematically derived algorithms.Yet,what we e in most papers is an emphasis on the mathematics and the algorithmic derivations,for which(at least for good papers),all necessary details are provided,so that the algorithm can be understood and reproduced without much difficulty.When it comes to the data or time ries to which the algorithms are applied,only nice pictures or generic statistics on the performance of the algorithm are provided,which are barely reproducable.
2Turn an art
Of cour,this lack of reproducibility basically originates in practical considerations,as one could not expect that papers would contain the complete datat,especially when it is huge.
A scientifically acceptable solution would be to make datats publically available onfloppys or CD-roms.Of cour,while rather expensive,this solution would also have its practical limitations of compatibility of data formats between different measurement and computing environments.
It goes without saying that the World Wide Web can contribute significantly to solve the reproducibilit
y problem hinted at in the introductory ction.We propo to construct a web-site,which we have called DAISY,which stands for Da taba for I dentification of Sy stems. The key idea is that authors,having published a paper on system identification or signal processing,submit the datat that was ud as an illustration,to DAISY,hence making it publically available1.
The best way to get acquainted with DAISY is to consult it at its World Wide Web URL: www.esat.kuleuven.ac.be/sista/daisy
The central objects in DAISY are datats,which,once submitted,undergo a(moderate) review procedure(tofilter out’impossible’or low quality datats)and,when accepted,are publically available on the Web2.Datats are grouped according to data categories,which at the time of writing consist of process industry hane-ethylene destillation column,glass furnace,...),electrical systems,mechanical wingflutter data, CD-player arm data,...),biomedical Fetal ECG measurements,...),biochem-ical systems,econometric data,environmental systems,’classical’datats,thermal datats.
1The issue of reproducibility requires that we agree on how to refer to DAISY in papers that will u some of its datats.We propo the following reference:De Moor B.(ed.).DAISY:Databa for the Id
团队口号8字押韵entification of Systems,Dept.of Electrical Engineering,ESAT/SISTA,K.U.Leuven,Belgium, URL:www.esat.kuleuven.ac.be/sista/daisy,+date of visit,name of datat,name of ction and code number.
2We take it for granted that all submitted datats have been cleared from any confidentiality agreement between the owner of the system on which the data were obtained and the person and/or organization that submitted the datat to DAISY.
There is an automatic submission procedure in which some characteristic parameters of the datat need to be described(sampling frequency,number of data,number of inputs and outputs and their units,)(e the website for details).Also available are an extended bibliography of more than100books on system identification and signal process-ing,a survey with World Wide Web hyperlinks to existing software packages and existing databas of datats on the Web.
3Making it work:L’app´e tit vient en mangeant!
While providing a basic solution to the problem of reproducibility in system identification, we achieve other benefits as well:Some of the datats in DAISY will evolve in due time into real benchmarks3,that will facilitate a comparison of the performance of algorithms. More generally,DAIS
Y can become instrumental in establishing comparisons of concepts, methods and algorithms.One and the same datat could be ud to asss the quality of veral variations of the same method,or,more generally,to compare the performance of different methods derived in different ’classical’system identification, prediction error methods,subspace methods,maximum likelihood methods,structured to-tal least squares,time versus frequency domain approaches,linear versus nonlinear,neural, )or different software environments(like Matlab,).DAISY will also stimulate collaboration and interaction between rearchers,organizations and compa-nies active in system identification.In particular,such a collaboration might enhance the cost-effectiveness of experiments,since measurement t-ups will not have to be repeated. DAISY will also be instrumental in providing inspiration to people in industry,when they e how certain datats are reminiscent to the application they have in mind.And why not u DAISY as a didactical tool,by inviting students to apply to real datats,in their home-works,the methods taught in system identification cours.Last but not least,a datat submitted to DAISY,will,on the average,be longer available in time than it would be with the author,who might have decided to move to another address or started a career in another domain than system identification.While experimental t-ups cea to exist at a certain moment in time,the datats that were obtained from it,will remain available under DAISY.
4Conclusions
We have been describing how an important deficiency in thefield of system identification, namely the reproducibility of datats and time ries,can be cured via DAISY,a Databa for Identification of Systems.We would like to invite all rearchers active in system iden-tification,data and time ries analysis to submit datats and provide us with feedback, suggestions for DAISY can be consulted at
www.esat.kuleuven.ac.be/sista/daisy
3All access to datats in DAISY are logged.A datat will evolve into a benchmark when it becomes a leader in the hitting statistics.The can be consulted in DAISY by just hitting a push button.
