Spatial methods: From spatial statistics through spatial econometrics to spatial machine learning

Katarzyna

Kopczewska

Associate professor at Faculty of Economic Sciences, University of Warsaw, Poland

An expert in quantitative methods applied to economics, she develops a methodology of spatial statistics, spatial econometrics and spatial machine learning. She conducts research on spatial economic and social phenomena such as business location, house prices or spatial distribution of population. She has published more than 20 papers in high-impact journals (e.g. Scandinavian Journal of Statistics, Spatial Statistics, Spatial Economic Analysis, Papers in Regional Science, Annals of Regional Science, Economic Modelling, PlosONE, Land Use Policy), co-authored 9 books (e.g. in Routledge, Edward Elgar, Springer Nature) and many other publications.

She is a member of the EOC of the European Regional Science Association (ERSA) and Councillor-at-Large of the Regional Science Association International (RSAI), and served as secretary of ERSA’s EPAINOS award. She is a co-editor of the journals Regional Science Policy and Practice and Networks and Spatial Economics. She is based at the University of Warsaw, where she was Dean of Students for eight years. She teaches in the Data Science MA, Computer Science and Econometrics MA and PhD programmes.

Workshop description

This workshop has three parts, each dealing with different approaches to spatial modelling. It addresses the challenges that arise with low granularity spatial data (geolocated points, grids, pixels, rasters). Social data (population, business location) will be used as an example, with the possibility of extending the methods to broader applications. The workshops will focus on the design and use of quantitative concepts and will be accompanied by R software codes.

Day 1

State of the art and challenges in spatial analysis: how new data sources are forcing progress in methodology

Day 2

Spatial statistics: what’s beyond Moran’s I? Solutions based on entropy, rings, kernels, nearest neighbours and unsupervised learning to deal with spatial density, heterogeneity, autocorrelation and agglomeration.

Day 3

Spatial econometrics: are we condemned to the spatial weight matrix? Alternative approaches using local coefficients, radial neighbourhood data and machine learning models.

The workshop is aimed at researchers at all stages of their career who need a methodological and practical overview of what is currently happening in the field of spatial quantitative methods. The workshops do not require fluency in mathematical notation and R programming, although analytical skills are highly desirable.

VENUE

Computer room 34

Faculty of Economics & Business

Trg J. F. Kennedyja 6, Zagreb


DATES

June 11-13, 2024

Tuesday, Wednesday & Thursday


TIME

Every day from 16:30 to 19:00

Workshop Registration

Klub R

  • S obzirom na izuzetan rast korisnika R-a i njegove statističke primjene, kako u akademskoj zajednici tako i u praksi, odlučili smo osnovati Klub R za članove HSD-a koji bi razmjenjivali i širili stečena znanja o upotrebi R-a te organizirali radionice i stručne susrete, što je osnovni cilj tog kluba.
  • Misija Kluba R je promoviranje i poticanje upotrebe R programskog okruženja te pružanje pomoći i podrške svojim članovima.
  • Vizija Kluba R je podići razinu pismenosti u R programskom okruženju.

R programski jezik

  • R je programski jezik slobodnog dohvata koji nudi široki izbor metoda za statističku i grafičku analizu podataka, a glavne su mu prednosti što se jako inuitivni kodovi (naredbe) mogu jednostavno nadograđivati i/ili prilagođavati potrebama korisnika, javno dostupni pimjeri mogu se u potpunosti reproducirati, a rad s kompleksnim objektima lako se autimatizira.
  • Djeluje na raznim UNIX platformama, Windows i MacOS. Preporuča se instalirati i RStudio kao integrirano razvojno okruženje (IDE) koje olakšava rad s R-om jer uključuje mnoge korisne funkcionalnosti, kao što su interaktivna konzola, alati za vizualizaciju, prozori za pisanje skripti, pregledavanje radnog prostora, praćenje povijesti naredbi, instaliranih paketa, itd. S prvom instalacijom R-a dobivaju se osnovni paketi koji podržavaju sam programski jezik.

Instalacija

R instalacija pomću poveznice 

RStudio (Desktop) instalacija pomoću poveznice 

Možete direktno pristupiti R programskom okruženju u vašem web pregledniku, bez potrebe za instalacijom R-a i RStudia pomoću platforme Posit Cloud. Navedeni web servis dostupan je na poveznici

RStudio postaje Posit (novi brand s proširenjem za Python) 2022. godine.

 

JOSIP ARNERIĆ

Voditelj “Kluba R”