Analyzing meta-datasets at an instance-level

• Grant number: 24/07655-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2024
• Status: Discontinued
• Field of knowledge: Physical Sciences and Mathematics - Computer Science
• Principal Investigator: Ana Carolina Lorena
• Grantee: Diogo Bueno Rodrigues
• 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
• Associated scholarship(s): 24/16562-2 - Determining empirical bounds of a dataset hardness embedding using the Instance Space Analysis framework, BE.EP.IC

Abstract

Meta-learning (MtL) has traditionally been focused in the analysis of a pool of datasets and how their characteristics relate to Machine Learning (ML) classification performance. But a more fine-grained analysis can be done at the instance-level, where particular characteristics of each instance of a dataset are regarded instead. Building on previous work of the research group, this scientific initiation project will study and validate MtL studies at an instance-level and understand how they can contribute to identify potential data quality issues inside a dataset.