A Comparative Study on Morphological Neural Networks for Binary Classification

• Luana Felipe de Barros UNICAMP
• Marcos Eduardo Valle UNICAMP

DOI: https://doi.org/10.5753/eniac.2021.18292

Abstract

Morphological neural networks represent a class of artificial neural networks whose neurons perform an operation from mathematical morphology followed by the application of an activation function. This paper provides a comparative study of different approaches that use morphological neural networks. Specifically, according to the training rule, we review incremental approaches, approaches based on maximum descent methods, extreme learning machines, and convex-concave optimization procedures. Computational experiments showed that, on average, the reduced dilation-erosion perceptron with bagging and ensemble strategies had better results in several binary classification problems.

References

Arce, F., Zamora, E., Sossa, H., and Barrón, R. (2018). Differential evolution training algorithm for dendrite morphological neural networks. Applied Soft Computing, 68:303–313.

Charisopoulos, V. and Maragos, P. (2017). Morphological perceptrons: Geometry and training algorithms. In Angulo, J., Velasco-Forero, S., and Meyer, F., editors, Mathematical Morphology and Its Applications to Signal and Image Processing, pages 3–15, Cham. Springer International Publishing.

de A. Araújo, R. (2011). A class of hybrid morphological perceptrons with application in time series forecasting. Knowledge-Based Systems, 24(4):513 – 529.

Franchi, G., Fehri, A., and Yao, A. (2020). Deep morphological networks. Pattern Recognition, 102:107246.

Haykin, S. (2009). Neural Networks and Learning Machines. Prentice-Hall, Upper Saddle River, NJ, 3rd edition edition.

Hernández, G., Zamora, E., Sossa, H., Téllez, G., and Furlán, F. (2019). Hybrid neural networks for big data classification. Neurocomputing, page S0925231219314560.

Huang, G., Huang, G. B., Song, S., and You, K. (2015). Trends in extreme learning machines: A review.

Mondal, R., Santra, S., and Chanda, B. (2019). Dense morphological network: An universal function approximator.

Pessoa, L. and Maragos, P. (2000). Neural networks with hybrid morphological/rank/linear nodes: a unifying framework with applications to handwritten character recognition. Pattern Recognition, 33:945–960.

Ritter, G. X. and Sussner, P. (1996). An Introduction to MorphoIn Proceedings of the 13th International Conference on logical Neural Networks. Pattern Recognition, pages 709–717, Vienna, Austria.

Ritter, G. X. and Urcid, G. (2003). Lattice Algebra Approach to Single-Neuron Computation. IEEE Transactions on Neural Networks, 14(2):282–295.

Sussner, P. (1998). Morphological perceptron learning. In IEEE International Symposium on Intelligent Control Proceedings, pages 477–482. IEEE.

Sussner, P. and Campiotti, I. (2020). Extreme learning machine for a new hybrid morphological/linear perceptron. Neural Networks, 123:288 – 298.

Sussner, P. and Esmi, E. L. (2011). Morphological Perceptrons with Competitive Learning: Lattice-Theoretical Framework and Constructive Learning Algorithm. Information Sciences, 181(10):1929–1950.

Valle, M. E. (2020). Reduced dilation-erosion perceptron for binary classification. Mathematics, 8(4):512.