Features optimization selection in hidden layers of deep learning based on graph clustering

EURASIP Journal on Wireless Communications and Networking

Research

Yifei Yuan¹  Hefei Gao²  Wei Wang² 

[1] The State Information Center (Administration Center of China E-government Network), 100045, Beijing, China;Tianjin Key Laboratory of Wireless Mobile Communications and Power Transmission, 300387, Tianjin, China;

关键词: Feature redundancy;    Graph cutting;    Graph neural network;    Hidden layers;    Spectral clustering;   

DOI  :  10.1186/s13638-023-02292-x

 received in 2023-04-24, accepted in 2023-08-14,  发布年份 2023

来源: Springer

【 摘 要 】

As it is widely known, big data can comprehensively describe the inherent laws governing various phenomena. However, the effective and efficient analysis of available data has become a major challenge in the fields of artificial intelligence, machine learning, data mining, and others. Deep learning, with its powerful learning ability and effective data-processing methods, has been extensively researched and applied in numerous academic domains. Nevertheless, the data obtained during the deep learning process often exhibits feature homogenization, resulting in highly redundant features in the hidden layers, which, in turn, affects the learning process. Therefore, this paper proposes an algorithm based on graph clustering to optimize the features of hidden layer units, with the aim of eliminating redundancy and improving learner generation.

【 授权许可】

CC BY   
© Springer Nature Switzerland AG 2023

【 图 表 】

12888_2023_5012_Article_IEq5.gif

Fig. 7

12888_2023_5012_Article_IEq3.gif

Fig. 1

12888_2023_5012_Article_IEq2.gif

12888_2023_5113_Article_IEq7.gif

12888_2023_5113_Article_IEq6.gif

12888_2023_5113_Article_IEq5.gif

12888_2023_5113_Article_IEq4.gif

12888_2023_5113_Article_IEq2.gif

12888_2023_5113_Article_IEq1.gif

Fig. 4

Fig. 1

Fig. 3

Fig. 1

Fig. 2

40517_2023_261_Article_IEq24.gif

40517_2023_261_Article_IEq45.gif

Fig. 3

40517_2023_261_Article_IEq43.gif

Fig. 2

Fig. 2

40517_2023_261_Article_IEq40.gif

Fig. 1

Fig. 1

40517_2023_261_Article_IEq37.gif

40517_2023_261_Article_IEq36.gif

40517_2023_261_Article_IEq35.gif

40517_2023_261_Article_IEq34.gif

40517_2023_261_Article_IEq31.gif

40517_2023_261_Article_IEq30.gif

Fig. 1

Fig. 1

13690_2023_1154_Article_IEq16.gif

Fig. 1

13690_2023_1154_Article_IEq14.gif

13690_2023_1154_Article_IEq13.gif

40517_2023_261_Article_IEq17.gif

13690_2023_1154_Article_IEq12.gif

13690_2023_1154_Article_IEq18.gif

13690_2023_1154_Article_IEq11.gif

13690_2023_1154_Article_IEq10.gif

13690_2023_1154_Article_IEq17.gif

13690_2023_1154_Article_IEq8.gif

13690_2023_1154_Article_IEq7.gif

13690_2023_1154_Article_IEq6.gif

13690_2023_1154_Article_IEq5.gif

40517_2023_261_Article_IEq6.gif

40517_2023_261_Article_IEq5.gif

40517_2023_261_Article_IEq4.gif

40517_2023_261_Article_IEq2.gif

Fig. 4

Fig. 1

13690_2023_1154_Article_IEq2.gif

Fig. 1

Fig. 1

Fig. 5

Fig. 1

Fig. 2

Fig. 3

Fig. 9

Fig. 3

Fig. 2

Fig. 8

12974_2023_2864_Article_IEq1.gif

Fig. 1

Fig. 4

Fig. 1

Fig. 1

13690_2023_1172_Article_IEq1.gif

42490_2023_74_Article_IEq17.gif

42490_2023_74_Article_IEq13.gif

42490_2023_74_Article_IEq12.gif

Fig. 2

Fig. 7

Fig. 1

Fig. 6

Fig. 1

Fig. 6

42490_2023_74_Article_IEq3.gif

Fig. 5

Fig. 1

Fig. 3

Fig. 3

Fig. 1

Fig. 5

40798_2023_622_Article_IEq15.gif

Scheme 1

40798_2023_622_Article_IEq7.gif

Fig. 5

13750_2023_309_Article_IEq64.gif

13690_2023_1151_Article_IEq9.gif