Data-Centric Approaches to Leveraging Large Language Models for Missing Data Imputation

• Grant number: 24/23791-8
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: July 01, 2025
• End date: February 28, 2029
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Ana Carolina Lorena
• Grantee: Arthur Dantas Mangussi
• Host Institution: Divisão de Ciência da Computação (IEC). Instituto Tecnológico de Aeronáutica (ITA). Ministério da Defesa (Brasil). São José dos Campos , SP, Brazil
• Associated research grant: 21/06870-3 - Beyond algorithm selection: meta-learning for data and algorithm analysis and understanding, AP.JP2

Abstract

Real-world data often present challenges such as imbalance, noise, and missing values. Recently, the Machine Learning literature has demonstrated that improving model performance also requires enhancing the quality of the data used for training, giving rise to a new research field known as Data-Centric Artificial Intelligence. Missing values, defined as the absence of information in one or more columns of a dataset, represent a common challenge in this context. The literature proposes various strategies to impute these values, ranging from simpler methods to advanced models based on Deep Learning. With the recent advancements in Generative Artificial Intelligence and Large Language Models (LLMs), new investigations have begun exploring the use of LLMs for missing value imputation. Therefore, this doctoral research proposal aims to investigate methodologies based on LLMs to perform missing data imputation efficiently and robustly from the perspective of Data-Centric AI.