Econometrics of Networks with Machine Learning

Oliver Kiss, Gyorgy Ruzicska

    Research output: Contribution to Book/Report typesChapterpeer-review

    Abstract (may include machine translation)

    Graph structured data, called networks, can represent many economic activities and phenomena. Such representations are not only powerful for developing economic theory but are also helpful in examining their applications in empirical analyses. This has been particularly the case recently as data associated with networks are often readily available. While researchers may have access to real-world network structured data, in many cases, their volume and complexities make analysis using traditional econometric methodology prohibitive. One plausible solution is to embed recent advancements in computer science, especially machine learning algorithms, into the existing econometric methodology that incorporates large networks. This chapter aims to cover a range of examples where existing algorithms in the computer science literature, machine learning tools, and econometric practices can complement each other. The first part of the chapter provides an overview of the challenges associated with high-dimensional, complex network data. It discusses ways to overcome them by using algorithms developed in computer science and econometrics. The second part of this chapter shows the usefulness of some machine learning algorithms in complementing traditional econometric techniques by providing empirical applications in spatial econometrics.

    Original languageEnglish
    Title of host publicationEconometrics with Machine Learning
    EditorsFelix Chan, László Mátyás
    PublisherSpringer Science and Business Media Deutschland GmbH
    Pages177-215
    Number of pages39
    ISBN (Electronic)978-3-031-15149-1
    ISBN (Print)978-3-031-15148-4, 978-3-031-15151-4
    DOIs
    StatePublished - 2022

    Publication series

    NameAdvanced Studies in Theoretical and Applied Econometrics
    Volume53
    ISSN (Print)1570-5811
    ISSN (Electronic)2214-7977

    Fingerprint

    Dive into the research topics of 'Econometrics of Networks with Machine Learning'. Together they form a unique fingerprint.

    Cite this