Título:	Algoritmo versus humanos : a assertividade e comportamento da racionalidade e intuição no processo de tomada de decisão
Autor(es):	Pereira, Diego Vinícius de Castro
Orientador(es):	Reyes Júnior, Edgar
Assunto:	Processo decisório Racionalidade Aprendizagem de máquina Decisões intuitivas
Data de publicação:	31-Ago-2022
Data de defesa:	30-Jul-2021
Referência:	PEREIRA, Diego Vinícius de Castro. Algoritmo versus humanos: a assertividade e comportamento da racionalidade e intuição no processo de tomada de decisão. 2021. 287 f., il. Tese (Doutorado em Administração) — Universidade de Brasília, Brasília, 2021.
Abstract:	The decision-making perspective has been more elaborated in normative research, which prescribes how decisions should be made. This type of research defines the criteria for a rational decision. However, total rationality could only be achieved by mathematical formulas or computer programs. On the other hand, the use of intuition is an important aspect of the decision-making process. But the literature presents an apparent dichotomy, with authors defending the use of intuition and defending rationality in decision making. In view of this dichotomy, the present work aims to investigate rational and intuitive decisions. For the rational decision, machine learning algorithms were used, a subfield of artificial intelligence. Thus, a model was developed for the recommendation and classification of potential candidates for Postdoctoral fellowships abroad. Initially, a pre-selection of the attributes was carried out, and 51 variables were computed at the end of the pre-processing. The training of the algorithm was carried out by means of 887 lattes curricula of post-doctorates abroad, in which four groups were identified. The groups were named according to their characteristics, with the following order of relevance: Score, Jobs_Orientations, IES_baixo_rank and Researcher_Senior. For the intuitive decision, a quasi-experiment was carried out composed of two experimental groups (Ad Hoc Evaluators and Doctors) and a control group (Students). The participants in each group were later classified as intuitive or rational. After analyzing the data, the results were presented in three different scenarios, and in the high validity scenario, the algorithm showed greater assertiveness in relation to humans, although intuitive specialists also showed significant assertiveness. In turn, in the uncertain scenario, which presents the internationalization variables, the algorithm showed assertiveness inferior to that of humans and again the intuitive specialists stood out in relation to the other groups. Finally, in the scenario that mixes uncertainties and validity, the algorithm and the intuitive specialists presented similar assertions in their decisions. It is noteworthy that the accuracy and recall techniques were used to verify the assertiveness and it consists in correctly recommending candidates before and after conducting the postdoctoral study abroad. In addition, through the main variables used for the decision, it was identified that the algorithm behaved more similarly to the intuitive specialist. Thus, this work presents as a theoretical contribution that the dichotomy between rational and intuition may be false. Rational and intuitive expert decisions can deliver good results and both can be relevant in a decision scenario. In addition, this work demonstrated that the computational tools are limited and that the algorithm used for the rational decision presented a behavior more similar to an intuitive human specialist, therefore, this behavior more similar to an intuitive specialist in machine learning algorithms and the possibility processing of different data formats can affect the rationality of decisions. Finally, this paper suggests the ideal candidate profile to be awarded postdoctoral scholarships abroad, the most suitable being those with a low number of publications, but with a high impact. In addition, these ideal candidates presented a lower number of orientations, low presentation of papers and participation in events, that is, the focus of these researchers was only high impact on their publications. Regarding the most suitable for Ad Hoc evaluators, the intuitive specialist profile was identified as the most assertive.
Unidade Acadêmica:	Faculdade de Economia, Administração, Contabilidade e Gestão de Políticas Públicas (FACE) Departamento de Administração (FACE ADM)
Informações adicionais:	Tese (doutorado) — Universidade de Brasília, Faculdade de Economia, Administração e Contabilidade e Gestão Pública, Programa de Pós-Graduação em Administração, 2021.
Programa de pós-graduação:	Programa de Pós-Graduação em Administração
Licença:	A concessão da licença deste item refere-se ao termo de autorização impresso assinado pelo autor com as seguintes condições: Na qualidade de titular dos direitos de autor da publicação, autorizo a Universidade de Brasília e o IBICT a disponibilizar por meio dos sites www.bce.unb.br, www.ibict.br, http://hercules.vtls.com/cgi-bin/ndltd/chameleon?lng=pt&skin=ndltd sem ressarcimento dos direitos autorais, de acordo com a Lei nº 9610/98, o texto integral da obra disponibilizada, conforme permissões assinaladas, para fins de leitura, impressão e/ou download, a título de divulgação da produção científica brasileira, a partir desta data.
Aparece nas coleções:	Teses, dissertações e produtos pós-doutorado

Os itens no repositório estão protegidos por copyright, com todos os direitos reservados, salvo quando é indicado o contrário.