Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples
In recent years, self-paced learning (SPL) has attracted much attention due to its improvement to nonconvex optimization based machine learning algorithms. As a methodology introduced from human learning, SPL dynamically evaluates the learning difficulty of each sample and provides the weighted lear...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2019-01-01
|
| Series: | Complexity |
| Online Access: | http://dx.doi.org/10.1155/2019/8127869 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832549311628443648 |
|---|---|
| author | Qi Zhu Ning Yuan Donghai Guan |
| author_facet | Qi Zhu Ning Yuan Donghai Guan |
| author_sort | Qi Zhu |
| collection | DOAJ |
| description | In recent years, self-paced learning (SPL) has attracted much attention due to its improvement to nonconvex optimization based machine learning algorithms. As a methodology introduced from human learning, SPL dynamically evaluates the learning difficulty of each sample and provides the weighted learning model against the negative effects from hard-learning samples. In this study, we proposed a cognitive driven SPL method, i.e., retrospective robust self-paced learning (R2SPL), which is inspired by the following two issues in human learning process: the misclassified samples are more impressive in upcoming learning, and the model of the follow-up learning process based on large number of samples can be used to reduce the risk of poor generalization in initial learning phase. We simultaneously estimated the degrees of learning-difficulty and misclassified in each step of SPL and proposed a framework to construct multilevel SPL for improving the robustness of the initial learning phase of SPL. The proposed method can be viewed as a multilayer model and the output of the previous layer can guide constructing robust initialization model of the next layer. The experimental results show that the R2SPL outperforms the conventional self-paced learning models in classification task. |
| format | Article |
| id | doaj-art-7c1405ab307a4603ac6eb9aa4d064b3b |
| institution | Kabale University |
| issn | 1076-2787 1099-0526 |
| language | English |
| publishDate | 2019-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Complexity |
| spelling | doaj-art-7c1405ab307a4603ac6eb9aa4d064b3b2025-02-03T06:11:32ZengWileyComplexity1076-27871099-05262019-01-01201910.1155/2019/81278698127869Cognitive Driven Multilayer Self-Paced Learning with Misclassified SamplesQi Zhu0Ning Yuan1Donghai Guan2College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaCollege of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaIn recent years, self-paced learning (SPL) has attracted much attention due to its improvement to nonconvex optimization based machine learning algorithms. As a methodology introduced from human learning, SPL dynamically evaluates the learning difficulty of each sample and provides the weighted learning model against the negative effects from hard-learning samples. In this study, we proposed a cognitive driven SPL method, i.e., retrospective robust self-paced learning (R2SPL), which is inspired by the following two issues in human learning process: the misclassified samples are more impressive in upcoming learning, and the model of the follow-up learning process based on large number of samples can be used to reduce the risk of poor generalization in initial learning phase. We simultaneously estimated the degrees of learning-difficulty and misclassified in each step of SPL and proposed a framework to construct multilevel SPL for improving the robustness of the initial learning phase of SPL. The proposed method can be viewed as a multilayer model and the output of the previous layer can guide constructing robust initialization model of the next layer. The experimental results show that the R2SPL outperforms the conventional self-paced learning models in classification task.http://dx.doi.org/10.1155/2019/8127869 |
| spellingShingle | Qi Zhu Ning Yuan Donghai Guan Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples Complexity |
| title | Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples |
| title_full | Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples |
| title_fullStr | Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples |
| title_full_unstemmed | Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples |
| title_short | Cognitive Driven Multilayer Self-Paced Learning with Misclassified Samples |
| title_sort | cognitive driven multilayer self paced learning with misclassified samples |
| url | http://dx.doi.org/10.1155/2019/8127869 |
| work_keys_str_mv | AT qizhu cognitivedrivenmultilayerselfpacedlearningwithmisclassifiedsamples AT ningyuan cognitivedrivenmultilayerselfpacedlearningwithmisclassifiedsamples AT donghaiguan cognitivedrivenmultilayerselfpacedlearningwithmisclassifiedsamples |