Banca de DEFESA: Cláudia Raquel da Rocha Eirado
Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.STUDENT : Cláudia Raquel da Rocha Eirado
DATE: 07/03/2025
TIME: 14:00
LOCAL: Vídeo Conferência
TITLE:
Essays on economics using networked data sets and machine learning.
KEY WORDS:
#centrality, #complexnetworks, #indirectagion, #machinelearning, #artificial intelligence, #systemic risk #digital twins
PAGES: 85
BIG AREA: Ciências Sociais Aplicadas
AREA: Economia
SUMMARY:
This work comprises three articles in Economics that utilize network data sets and machine learning. In the first article, we conduct a review of machine learning applications to solve complex network problems. We cover concepts of machine learning, including supervised learning, unsupervised learning, and reinforcement learning, along with methods such as clustering, embedding, and PCA (Principal Component Analysis). Additionally, we explore network construction and centrality concepts, addressing node and link prediction. The article also discusses natural language approaches, incorporating theories from Natural Language Processing (NLP). The second article investigates the concept of systemic risk in the financial domain, exploring its potential to trigger indirect contagion. A fundamental part of the research involves applying a model that uses a news similarity network to predict stationary probabilities as a proxy for network centrality and relationships between companies. The study establishes connections among companies, identifying pathways of indirect contagion. By analyzing interactions and the spread of contagion between companies based on news articles, the study seeks to uncover insights into interconnectivity and cascading effects within the financial system, as well as potential impacts on other sectors. The article concludes with a discussion of the potential applications of AI and ML in understanding and predicting systemic risk in the financial landscape. The third article presents an empirical exercise on Digital Twin Modeling applied to the EU Emissions Trading System (EU ETS). We use EU ETS transaction data to identify patterns of interconnection between countries. To achieve this, we build complex networks to outline relationships among nations, representing contagion pathways. Using Digital Twins, we simulate the entry and exit of new agents and the formation of new connections based on predictive analysis through machine learning models.
COMMITTEE MEMBERS:
Interno - 1196877 - BERNARDO PINHEIRO MACHADO MUELLER
Presidente - 1642911 - DANIEL OLIVEIRA CAJUEIRO
Externo à Instituição - Douglas Sad Silveira - UFJF
Externo ao Programa - 1997721 - HERBERT KIMURA - nullExterno à Instituição - SAULO BENCHIMOL BASTOS - BACEN