A. M. Carvalho

Universidade Federal do ABC

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Ronaldo C. Prati

Universidade Federal do ABC

Michelle Chaves Kuroda

Universidade Estadual de Campinas (UNICAMP)

Alexandre Campane Vidal

Universidade Estadual de Campinas (UNICAMP)

Cooperative Institutions

Universidade Estadual de Campinas (UNICAMP)

Hospital de Clínicas da Unicamp

Universidade Federal do ABC

Centro Universitário Fundação Santo André

Universidade de São Paulo

Brazilian Society of Computational and Applied Mathematics

Asociación Española de Urología

Pontifícia Universidade Católica de Campinas

Intel (United States)

Ibero American University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Improving kNN classification under Unbalanced Data. A New Geometric Oversampling Approach

2022 International Joint Conference on Neural Networks (IJCNN) (2018)

A. M. Carvalho Ronaldo C. Prati

Training classifiers with unbalanced data is one of the main challenges in the field of Machine Learning. Some techniques that try to get around this problem have been proposed, where one of most important is SMOTE, which artificially generates new instances by interpolating pairs of original instances. This paper proposes a new approach to the balancing of classifier training data. It is a geometric and spatial approach, that uses tetrahedralization by Delaunay tessellation to generate new artificial instances. This allow us to go beyond single pair interpolation. Applying this new method, we can notice an improvement in the classification quality (in terms of AUC) of the kNN classification algorithm when compared with SMOTE.

Oversampling

Interpolation

10.1109/ijcnn.2018.8489411

Cite

Citations (7)

DTO-SMOTE: Delaunay Tessellation Oversampling for Imbalanced Data Sets

Information (2020)

A. M. Carvalho Ronaldo C. Prati

One of the significant challenges in machine learning is the classification of imbalanced data. In many situations, standard classifiers cannot learn how to distinguish minority class examples from the others. Since many real problems are unbalanced, this problem has become very relevant and deeply studied today. This paper presents a new preprocessing method based on Delaunay tessellation and the preprocessing algorithm SMOTE (Synthetic Minority Over-sampling Technique), which we call DTO-SMOTE (Delaunay Tessellation Oversampling SMOTE). DTO-SMOTE constructs a mesh of simplices (in this paper, we use tetrahedrons) for creating synthetic examples. We compare results with five preprocessing algorithms (GEOMETRIC-SMOTE, SVM-SMOTE, SMOTE-BORDERLINE-1, SMOTE-BORDERLINE-2, and SMOTE), eight classification algorithms, and 61 binary-class data sets. For some classifiers, DTO-SMOTE has higher performance than others in terms of Area Under the ROC curve (AUC), Geometric Mean (GEO), and Generalized Index of Balanced Accuracy (IBA).

Oversampling

Tessellation (computer graphics)

10.3390/info11120557

Cite

Citations (10)

Classification of Stromatolites Using SOM

Proceedings (2011)

Michelle Chaves Kuroda A. M. Carvalho Alexandre Campane Vidal

The aim of this research is to analyze CT images of stromatolites originated from Salt Lake, RJ, Brazil, with a neural networks tool, called SOM, developed by Kohonen in the 80's. This research analyzed images of various sections of stromatolites and identified three classes: pore, sediment and rock. The automatically, quickly and with high accuracy are the advantages of this technique can be used as an alternative to conventional methods used, generating 3D images.

Self-organizing map

Salt lake

10.3997/2214-4609.20149712

Cite

Citations (0)

Classification of Well Using SOM and LVQ

Proceedings (2011)

Michelle Chaves Kuroda A. M. Carvalho Alexandre Campane Vidal

The objective of this research is to expand the possible applications of the tools of neural networks, SOM and LVQ, developed by Kohonen in the 80's, deepening the study of geological correlation between wells. This research analyzed values of well logs and found similar patterns that define layers between the wells analyzed. The automatically, quickly and with high accuracy are the advantages of this technique can be used as an alternative to conventional methods.

Self-organizing map

Learning vector quantization

10.3997/2214-4609.20149708

Cite

Citations (0)