TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Progressive Operational Perceptrons with Memory

Tutkimustuotosvertaisarvioitu

Standard

Progressive Operational Perceptrons with Memory. / Tran, Dat Thanh; Kiranyaz, Serkan; Gabbouj, Moncef; Iosifidis, Alexandros.

julkaisussa: Neurocomputing, 31.10.2019.

Tutkimustuotosvertaisarvioitu

Harvard

APA

Vancouver

Author

Tran, Dat Thanh ; Kiranyaz, Serkan ; Gabbouj, Moncef ; Iosifidis, Alexandros. / Progressive Operational Perceptrons with Memory. Julkaisussa: Neurocomputing. 2019.

Bibtex - Lataa

@article{8a777941ab9b46328bbd56bad2f49423,
title = "Progressive Operational Perceptrons with Memory",
abstract = "Generalized Operational Perceptron (GOP) was proposed to generalize the linear neuron model used in the traditional Multilayer Perceptron (MLP) by mimicking the synaptic connections of biological neurons showing nonlinear neurochemical behaviours. Previously, Progressive Operational Perceptron (POP) was proposed to train a multilayer network of GOPs which is formed layer-wise in a progressive manner. While achieving superior learning performance over other types of networks, POP has a high computational complexity. In this work, we propose POPfast, an improved variant of POP that signicantly reduces the computational complexity of POP, thus accelerating the training time of GOP networks. In addition, we also propose major architectural modications of POPfast that can augment the progressive learning process of POP by incorporating an information preserving, linear projection path from the input to the output layer at each progressive step. The proposed extensions can be interpreted as a mechanism that provides direct information extracted from the previously learned layers to the network, hence the term “memory”. This allows the network to learn deeper architectures and better data representations. An extensive set of experiments in human action, object, facial identity and scene recognition problems demonstrates that the proposed algorithms can train GOP networks much faster than POPs while achieving better performance compared to original POPs and other related algorithms.",
keywords = "Generalized operational perceptron, Progressive learning, Neural architecture learning",
author = "Tran, {Dat Thanh} and Serkan Kiranyaz and Moncef Gabbouj and Alexandros Iosifidis",
note = "EXT={"}Kiranyaz, Serkan{"} EXT={"}Iosifidis, Alexandros{"}",
year = "2019",
month = "10",
day = "31",
doi = "10.1016/j.neucom.2019.10.079",
language = "English",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier Science B.V.",

}

RIS (suitable for import to EndNote) - Lataa

TY - JOUR

T1 - Progressive Operational Perceptrons with Memory

AU - Tran, Dat Thanh

AU - Kiranyaz, Serkan

AU - Gabbouj, Moncef

AU - Iosifidis, Alexandros

N1 - EXT="Kiranyaz, Serkan" EXT="Iosifidis, Alexandros"

PY - 2019/10/31

Y1 - 2019/10/31

N2 - Generalized Operational Perceptron (GOP) was proposed to generalize the linear neuron model used in the traditional Multilayer Perceptron (MLP) by mimicking the synaptic connections of biological neurons showing nonlinear neurochemical behaviours. Previously, Progressive Operational Perceptron (POP) was proposed to train a multilayer network of GOPs which is formed layer-wise in a progressive manner. While achieving superior learning performance over other types of networks, POP has a high computational complexity. In this work, we propose POPfast, an improved variant of POP that signicantly reduces the computational complexity of POP, thus accelerating the training time of GOP networks. In addition, we also propose major architectural modications of POPfast that can augment the progressive learning process of POP by incorporating an information preserving, linear projection path from the input to the output layer at each progressive step. The proposed extensions can be interpreted as a mechanism that provides direct information extracted from the previously learned layers to the network, hence the term “memory”. This allows the network to learn deeper architectures and better data representations. An extensive set of experiments in human action, object, facial identity and scene recognition problems demonstrates that the proposed algorithms can train GOP networks much faster than POPs while achieving better performance compared to original POPs and other related algorithms.

AB - Generalized Operational Perceptron (GOP) was proposed to generalize the linear neuron model used in the traditional Multilayer Perceptron (MLP) by mimicking the synaptic connections of biological neurons showing nonlinear neurochemical behaviours. Previously, Progressive Operational Perceptron (POP) was proposed to train a multilayer network of GOPs which is formed layer-wise in a progressive manner. While achieving superior learning performance over other types of networks, POP has a high computational complexity. In this work, we propose POPfast, an improved variant of POP that signicantly reduces the computational complexity of POP, thus accelerating the training time of GOP networks. In addition, we also propose major architectural modications of POPfast that can augment the progressive learning process of POP by incorporating an information preserving, linear projection path from the input to the output layer at each progressive step. The proposed extensions can be interpreted as a mechanism that provides direct information extracted from the previously learned layers to the network, hence the term “memory”. This allows the network to learn deeper architectures and better data representations. An extensive set of experiments in human action, object, facial identity and scene recognition problems demonstrates that the proposed algorithms can train GOP networks much faster than POPs while achieving better performance compared to original POPs and other related algorithms.

KW - Generalized operational perceptron

KW - Progressive learning

KW - Neural architecture learning

U2 - 10.1016/j.neucom.2019.10.079

DO - 10.1016/j.neucom.2019.10.079

M3 - Article

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -