Marcelo de Sousa Balbino

Marcelo de Sousa Balbino

Auto-generated profile, based on publicly available information

Marcelo de Sousa Balbino

Institute of Exact Sciences, Pontifical Catholic University of Minas Gerais (PUC Minas), Belo Horizonte, Brazil

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 May 2023 in Expert Systems with Applications
Verified
Journal Article
Closed access
MBMarcelo de Sousa Balbino
Marcelo de Sousa Balbino
LZLuis E. Zárate
Luis E. Zárate
CNCristiane Neri Nobre
Cristiane Neri Nobre

15 March 2023 in Artificial Intelligence and Law
Verified
Journal Article
Closed access
PDF file
FOFábio M. Oliveira
Fábio M. Oliveira
MBMarcelo S. Balbino
Marcelo S. Balbino
LZLuis E. Zárate
Luis E. Zárate
FNFawn Ngo
Fawn Ngo
RGRamakrishna Govindu
Ramakrishna Govindu
CNCristiane Neri Nobre
Cristiane Neri Nobre

26 September 2023
Verified
Conference Paper
Open access
PDF file
OKOmar F. de P. e Krauss
Omar F. de P. e Krauss
MBMarcelo de S. Balbino
Marcelo de S. Balbino
CNCristiane Neri Nobre
Cristiane Neri Nobre
There is currently a growing concern about the explainability of machine learning algorithms. Explainability refers to the ability to understand and interpret the decisions made by the models, that is, the process by which a model arrives at a given prediction or classification. The counterfactual explanation involves creating alternative examples where the models prediction differs from the original. This work aims to raise and discuss essential features in the context of counterfactual explanation methods. For this, the CSSE and LORE methods will be evaluated and applied to twelve public databases, considering different characteristics regarding the number of attributes and data types. In this way, we can better understand their strengths and weaknesses using standardized metrics for different methods. This facilitates the selection and development of more effective strategies and helps to identify cases where one approach may outperform another regarding the quality of explanations. The survey measured the metrics validity, prolixity, sparsity, similarity, and hit rate. In general terms, the CSSE performed better in these metrics, except for sparsity. There is currently a growing concern about the explainability of machine learning algorithms. Explainability refers to the ability to understand and interpret the decisions made by the models, that is, the process by which a model arrives at a given prediction or classification. The counterfactual explanation involves creating alternative examples where the models prediction differs from the original. This work aims to raise and discuss essential features in the context of counterfactual explanation methods. For this, the CSSE and LORE methods will be evaluated and applied to twelve public databases, considering different characteristics regarding the number of attributes and data types. In this way, we can better understand their strengths and weaknesses using standardized metrics for different methods. This facilitates the selection and development of more effective strategies and helps to identify cases where one approach may outperform another regarding the quality of explanations. The survey measured the metrics validity, prolixity, sparsity, similarity, and hit rate. In general terms, the CSSE performed better in these metrics, except for sparsity.

27 August 2025 in Machine Learning
Verified
Journal Article
Closed access
PDF file
OKOmar F. P. Krauss
Omar F. P. Krauss
MBMarcelo S. Balbino
Marcelo S. Balbino
CNCristiane Neri Nobre
Cristiane Neri Nobre

2 October 2024 in Universal Access in the Information Society
Verified
Journal Article
Closed access
PDF file
CJCaroline R. S. Jandre
Caroline R. S. Jandre
AVAliene A. Villaça
Aliene A. Villaça
MBMarcelo S. Balbino
Marcelo S. Balbino
DMDebora De Miranda
Debora De Miranda
CNCristiane Neri Nobre
Cristiane Neri Nobre

28 November 2022
Verified
Conference Paper
Open access
PDF file
FOF. M. de Oliveira
F. M. de Oliveira
MBM. de S. Balbino
M. de S. Balbino
LZLuis E. Zárate
Luis E. Zárate
CNCristiane Neri Nobre
Cristiane Neri Nobre
Correctional institutions often develop rehabilitation programs to reduce the likelihood of inmates committing internal offenses and criminal recidivism after release. Therefore, it is necessary to identify the profile of each offender, both for the appropriate indication of a rehabilitation program and the level of internal security to which he must be submitted. In this context, this work aims to discover, from Machine Learning methods and the SHAP approach, which are the most significant characteristics in the prediction of misconduct by prisoners. For this, a database produced in 2004 through the Survey of Inmates in State and Federal Correctional Facilities in the United States of America, which provides nationally representative data on prisoners in state and federal facilities, was used. The predictive model based on Random Forest had the best performance; therefore, SHAP was applied to it to interpret the results. In addition, the attributes related to the type of crime committed, age at first arrest, drug use, mental or emotional health problems, having children, and being abused before arrest are more relevant in predicting internal misconduct. Thus, it is expected to contribute to the prior classification of an inmate, on time, use of programs and practices that aim to improve the lives of offenders, their reintegration into society, and, consequently, the reduction of criminal recidivism.