Gated recurrent units and temporal convolutional network for multilabel classification
Authors:
Loris Nanni,
Alessandra Lumini,
Alessandro Manfe,
Riccardo Rampon,
Sheryl Brahnam,
Giorgio Venturin
Abstract:
Multilabel learning tackles the problem of associating a sample with multiple class labels. This work proposes a new ensemble method for managing multilabel classification: the core of the proposed approach combines a set of gated recurrent units and temporal convolutional neural networks trained with variants of the Adam optimization approach. Multiple Adam variants, including novel one proposed…
▽ More
Multilabel learning tackles the problem of associating a sample with multiple class labels. This work proposes a new ensemble method for managing multilabel classification: the core of the proposed approach combines a set of gated recurrent units and temporal convolutional neural networks trained with variants of the Adam optimization approach. Multiple Adam variants, including novel one proposed here, are compared and tested; these variants are based on the difference between present and past gradients, with step size adjusted for each parameter. The proposed neural network approach is also combined with Incorporating Multiple Clustering Centers (IMCC), which further boosts classification performance. Multiple experiments on nine data sets representing a wide variety of multilabel tasks demonstrate the robustness of our best ensemble, which is shown to outperform the state-of-the-art. The MATLAB code for generating the best ensembles in the experimental section will be available at https://github.com/LorisNanni.
△ Less
Submitted 22 August, 2022; v1 submitted 8 October, 2021;
originally announced October 2021.
High performing ensemble of convolutional neural networks for insect pest image detection
Authors:
Loris Nanni,
Alessandro Manfe,
Gianluca Maguolo,
Alessandra Lumini,
Sheryl Brahnam
Abstract:
Pest infestation is a major cause of crop damage and lost revenues worldwide. Automatic identification of invasive insects would greatly speedup the identification of pests and expedite their removal. In this paper, we generate ensembles of CNNs based on different topologies (ResNet50, GoogleNet, ShuffleNet, MobileNetv2, and DenseNet201) altered by random selection from a simple set of data augmen…
▽ More
Pest infestation is a major cause of crop damage and lost revenues worldwide. Automatic identification of invasive insects would greatly speedup the identification of pests and expedite their removal. In this paper, we generate ensembles of CNNs based on different topologies (ResNet50, GoogleNet, ShuffleNet, MobileNetv2, and DenseNet201) altered by random selection from a simple set of data augmentation methods or optimized with different Adam variants for pest identification. Two new Adam algorithms for deep network optimization based on DGrad are proposed that introduce a scaling factor in the learning rate. Sets of the five CNNs that vary in either data augmentation or the type of Adam optimization were trained on both the Deng (SMALL) and the large IP102 pest data sets. Ensembles were compared and evaluated using three performance indicators. The best performing ensemble, which combined the CNNs using the different augmentation methods and the two new Adam variants proposed here, achieved state of the art on both insect data sets: 95.52% on Deng and 73.46% on IP102, a score on Deng that competed with human expert classifications. Additional tests were performed on data sets for medical imagery classification that further validated the robustness and power of the proposed Adam optimization variants. All MATLAB source code is available at https://github.com/LorisNanni/.
△ Less
Submitted 27 August, 2021;
originally announced August 2021.
