FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3200112955> ?p ?o ?g. }

• W3200112955 abstract "This dissertation is a collection of three essaysthat apply statistical methods to the study of internationalrelations. All three essays share the unifying theme of bridgingmachine learning and international relations. The first essayanalyzes the determinants of prospective countries joining theAsian Infrastructure Investment Bank (AIIB). I argue that lessdemocratic countries are more likely to join and that whenconsidering joining the institution countries learn from theirneighbors and from their own previous international organization(IO) interactions with China. Building on detailed panel data overthe institution's founding period, i.e. from October 24th, 2014 toMarch 31st, 2015, I fit both a probit model with time polynomialsand a Cox duration model to identify country characteristics thatcorrelate with joining the Asian Infrastructure Investment Bank. Ishow that countries with lower polity scores were more likely tojoin, that countries were also more likely to join after theirneighbors joined, and that the probability of joining was higherfor countries that had previously joined China-led IOs. Lastly, Ishow that countries under-represented in the existinginternational financial system were more likely to join. Myfindings highlight the importance of democracy in shaping themembership structure of the AIIB, demonstrate that countriesleverage information from their neighbors and from their previousinteractions with China to adjust the perceived risk of joining,and offer the first modern-day empirical support for the contestedmultilateralism framework. The second essay studies the performanceof gradient boosting machines (GBM) in predicting civil waronsets. GBM is a machine learning technique widely used both forclassification and regression tasks. With an eye on binaryclassification, I introduce the implementation of GBM and that ofregression trees, which GBM uses as a subroutine. I use MonteCarlo simulations to illustrate the intuition and details behindGBM's tree growing process. I also re-analyze an empirical example(Muchlinski et al., 2016) to compare GBM's performance with thatof random forest, using separation plot, area under the receiveroperating characteristic curve (ROC-AUC), nested cross-validationand area under the precision-recall curve (PR-AUC). The thirdessay proposes a measure of bilateral engagement using wordembedding. Engagement between two countries, whether it is betweenthe U.S. and Turkey, between China and Malaysia, or between Braziland India, is an often-discussed yet rarely measured concept. Iattempt to fill this gap by embedding countries on a yearly basisusing a large collection of news articles. With 1.86 million NewYork Times news articles published between 1987 and 2006, ameasure of dyadic engagement, valued between -1 and 1, is createdfor all existing countries and for all years, thus enablingdownstream analysis of, for example, civil war prediction, UNvoting buying, or Chinese FDI outflow. To demonstrate…" @default.
• W3200112955 created "2021-09-27" @default.
• W3200112955 creator A5007260596 @default.
• W3200112955 creator A5079594267 @default.
• W3200112955 date "2018-01-01" @default.
• W3200112955 modified "2023-09-24" @default.
• W3200112955 title "Essays on computational internationalrelations" @default.
• W3200112955 hasPublicationYear "2018" @default.
• W3200112955 type Work @default.
• W3200112955 sameAs 3200112955 @default.
• W3200112955 citedByCount "0" @default.
• W3200112955 crossrefType "journal-article" @default.
• W3200112955 hasAuthorship W3200112955A5007260596 @default.
• W3200112955 hasAuthorship W3200112955A5079594267 @default.
• W3200112955 hasConcept C162324750 @default.
• W3200112955 hasConcept C17744445 @default.
• W3200112955 hasConcept C191935318 @default.
• W3200112955 hasConcept C199539241 @default.
• W3200112955 hasConcept C2779707719 @default.
• W3200112955 hasConcept C2780510313 @default.
• W3200112955 hasConcept C555826173 @default.
• W3200112955 hasConcept C94625758 @default.
• W3200112955 hasConceptScore W3200112955C162324750 @default.
• W3200112955 hasConceptScore W3200112955C17744445 @default.
• W3200112955 hasConceptScore W3200112955C191935318 @default.
• W3200112955 hasConceptScore W3200112955C199539241 @default.
• W3200112955 hasConceptScore W3200112955C2779707719 @default.
• W3200112955 hasConceptScore W3200112955C2780510313 @default.
• W3200112955 hasConceptScore W3200112955C555826173 @default.
• W3200112955 hasConceptScore W3200112955C94625758 @default.
• W3200112955 hasLocation W32001129551 @default.
• W3200112955 hasOpenAccess W3200112955 @default.
• W3200112955 hasPrimaryLocation W32001129551 @default.
• W3200112955 hasRelatedWork W1974800624 @default.
• W3200112955 hasRelatedWork W2025476615 @default.
• W3200112955 hasRelatedWork W2029310714 @default.
• W3200112955 hasRelatedWork W2263679075 @default.
• W3200112955 hasRelatedWork W2329660705 @default.
• W3200112955 hasRelatedWork W2608498415 @default.
• W3200112955 hasRelatedWork W2766354431 @default.
• W3200112955 hasRelatedWork W2767978858 @default.
• W3200112955 hasRelatedWork W2789785507 @default.
• W3200112955 hasRelatedWork W2906676345 @default.
• W3200112955 hasRelatedWork W2951096842 @default.
• W3200112955 hasRelatedWork W3008568436 @default.
• W3200112955 hasRelatedWork W3122001533 @default.
• W3200112955 hasRelatedWork W3145199371 @default.
• W3200112955 hasRelatedWork W3162014298 @default.
• W3200112955 hasRelatedWork W3170467985 @default.
• W3200112955 hasRelatedWork W3181255655 @default.
• W3200112955 hasRelatedWork W3199560546 @default.
• W3200112955 hasRelatedWork W40529113 @default.
• W3200112955 hasRelatedWork W2188619790 @default.
• W3200112955 isParatext "false" @default.
• W3200112955 isRetracted "false" @default.
• W3200112955 magId "3200112955" @default.
• W3200112955 workType "article" @default.