scholarly journals Differential Entropy Feature Signal Extraction Based on Activation Mode and Its Recognition in Convolutional Gated Recurrent Unit Network

2021 ◽  
Vol 8 ◽  
Author(s):  
Yongsheng Zhu ◽  
Qinghua Zhong

In brain-computer-interface (BCI) devices, signal acquisition via reducing the electrode channels can reduce the computational complexity of models and filter out the irrelevant noise. Differential entropy (DE) plays an important role in emotional components of signals, which can reflect the area activity differences. Therefore, to extract distinctive feature signals and improve the recognition accuracy based on feature signals, a method of DE feature signal recognition based on a Convolutional Gated Recurrent Unit network was proposed in this paper. Firstly, the DE and power spectral density (PSD) of each original signal were mapped to two topographic maps, and the activated channels could be selected in activation modes. Secondly, according to the position of original electrodes, 1D feature signal sequences with four bands were reconstructed into a 3D feature signal matrix, and a radial basis function interpolation was used to fill in zero values. Then, the 3D feature signal matrices were fed into a 2D Convolutional Neural Network (2DCNN) for spatial feature extraction, and the 1D feature signal sequences were fed into a bidirectional Gated Recurrent Unit (BiGRU) network for temporal feature extraction. Finally, the spatial-temporal features were fused by a fully connected layer, and recognition experiments based on DE feature signals at the different time scales were carried out on a DEAP dataset. The experimental results showed that there were different activation modes at different time scales, and the reduction of the electrode channel could achieve a similar accuracy with all channels. The proposed method achieved 87.89% on arousal and 88.69% on valence.

Author(s):  
Joshua M. Epstein

This part describes the agent-based and computational model for Agent_Zero and demonstrates its capacity for generative minimalism. It first explains the replicability of the model before offering an interpretation of the model by imagining a guerilla war like Vietnam, Afghanistan, or Iraq, where events transpire on a 2-D population of contiguous yellow patches. Each patch is occupied by a single stationary indigenous agent, which has two possible states: inactive and active. The discussion then turns to Agent_Zero's affective component and an elementary type of bounded rationality, as well as its social component, with particular emphasis on disposition, action, and pseudocode. Computational parables are then presented, including a parable relating to the slaughter of innocents through dispositional contagion. This part also shows how the model can capture three spatially explicit examples in which affect and probability change on different time scales.


2021 ◽  
Vol 34 (1) ◽  
Author(s):  
Weixin Xu ◽  
Huihui Miao ◽  
Zhibin Zhao ◽  
Jinxin Liu ◽  
Chuang Sun ◽  
...  

AbstractAs an integrated application of modern information technologies and artificial intelligence, Prognostic and Health Management (PHM) is important for machine health monitoring. Prediction of tool wear is one of the symbolic applications of PHM technology in modern manufacturing systems and industry. In this paper, a multi-scale Convolutional Gated Recurrent Unit network (MCGRU) is proposed to address raw sensory data for tool wear prediction. At the bottom of MCGRU, six parallel and independent branches with different kernel sizes are designed to form a multi-scale convolutional neural network, which augments the adaptability to features of different time scales. These features of different scales extracted from raw data are then fed into a Deep Gated Recurrent Unit network to capture long-term dependencies and learn significant representations. At the top of the MCGRU, a fully connected layer and a regression layer are built for cutting tool wear prediction. Two case studies are performed to verify the capability and effectiveness of the proposed MCGRU network and results show that MCGRU outperforms several state-of-the-art baseline models.


2021 ◽  
Vol 98 ◽  
pp. 105254 ◽  
Author(s):  
Christian Urom ◽  
Hela Mzoughi ◽  
Ilyes Abid ◽  
Mariem Brahim

2005 ◽  
Vol 56 (4) ◽  
pp. 1049-1061 ◽  
Author(s):  
Richard A. Stein ◽  
Shuang Deng ◽  
N. Patrick Higgins

2021 ◽  
Vol 125 ◽  
pp. 107582 ◽  
Author(s):  
Yibo Wang ◽  
Pan Liu ◽  
Ming Dou ◽  
He Li ◽  
Bo Ming ◽  
...  

2010 ◽  
Vol 6 (S276) ◽  
pp. 527-529
Author(s):  
Xavier Dumusque ◽  
Nuno C. Santos ◽  
Stéphane Udry ◽  
Cristophe Lovis ◽  
Xavier Bonfils

AbstractSpectrographs like HARPS can now reach a sub-ms−1 precision in radial-velocity (RV) (Pepe & Lovis 2008). At this level of accuracy, we start to be confronted with stellar noise produced by 3 different physical phenomena: oscillations, granulation phenomena (granulation, meso- and super-granulation) and activity. On solar type stars, these 3 types of perturbation can induce ms−1 RV variation, but on different time scales: 3 to 15 minutes for oscillations, 15 minutes to 1.5 days for granulation phenomena and 10 to 50 days for activity. The high precision observational strategy used on HARPS, 1 measure per night of 15 minutes, on 10 consecutive days each month, is optimized, due to a long exposure time, to average out the noise coming from oscillations (Dumusque et al. 2011a) but not to reduce the noise coming from granulation and activity (Dumusque et al. 2011a and Dumusque et al. 2011b). The smallest planets found with this strategy (Mayor et al. 2009) seems to be at the limit of the actual observational strategy and not at the limit of the instrumental precision. To be able to find Earth mass planets in the habitable zone of solar-type stars (200 days for a K0 dwarf), new observational strategies, averaging out simultaneously all type of stellar noise, are required.


Sign in / Sign up

Export Citation Format

Share Document