When people observe the actions of others, they naturally try to understand the underlying intentions. This behavior is called action understanding, and it has an important influence on mental development, language comprehension, and socialization. In this study, we used functional near-infrared spectroscopy (fNIRS) to obtain brain signals related to action understanding and then classified different intentions. Aiming to overcome the drawbacks of traditional multiclass classification methods of one-versus-rest (OVR) and one-versus-one (OVO), in this paper, we propose a new effective method to solve multiclass classification that is a combination of OVR and OVO. Compared with OVO, this new method effectively improved the accuracy of four-class classification from 25% to 48%.
27 July 2020 in Mathematical Problems in Engineering
Permutations with low differential uniformity, high algebraic degree, and high nonlinearity over can be used as the substitution boxes for many block ciphers. In this paper, several classes of low differential uniformity permutations are constructed based on the method of choosing two permutations over to get the desired permutations. The resulted low differential uniformity permutations have high algebraic degrees and nonlinearities simultaneously, which provide more choices for the substitution boxes. Moreover, some numerical examples are provided to show the efficacy of the theoretical results.
The financial domain naturally contains multiple different types of entities such as stocks, product categories, investment participants, intermediaries, and customers, and interactions between these entities. This paper introduces a Graph-based Collaborative Filtering Category Recommendation (GCFCR) system as a first step in modelling the financial domain as an inter-connected, heterogeneous, dynamic system of nodes and edges. The goal of this paper is to identify customer interest based on the neighborhood of each node and make personalized suggestions or identify relevant content for each customer. Matching relevant products and services to customers is a key foundation of building and maintaining strong customer relationships, facilitating more personalized marketing which can ultimately result in increased customer activity, trust, and revenue. In this paper, we run a set of experiments to compare different recommendation techniques, concluding that the proposed GCFCR approach outperforms in this real-life application.
Several studies have been provided the principal activated regions and locations during action observation. However, it remains unclear that the dynamic link and information flows in different visual perspectives during the action observation tasks. The aim of this study was investigating the effective connectivity of different visual perspectives during action observation tasks and classify the perspectives. Eleven regions were selected as the regions of interest (ROI) by meta-analysis using the term of "action observation" for this study. We employed weighted and directed (WD) brain networks by Generalized partial directed coherence (GPDC) on the fMRI BOLD time series of action observation tasks from different visual perspectives. Afterward, the paired sample t-test was implemented to find the alteration of brain networks between the two perspectives. According to the results, global efficiency, local efficiency and cluster coefficient were statistically significantly different between first-person perspective (1PP) and third-person perspective (3PP) in Supplementary motor cortex (SMA) and Right inferior parietal lobe (RIPL) and Left inferior parietal lobe (LIPL). These three regions which we considered may provide the additional support for the visuospatial transformation from 3PP to 1PP. WD brain network topological measures of SMA, RIPL and LIPL were also selected to classify the 1PP and 3PP using support vector machine (SVM) with the best accuracy close to 90%.
When people observe the actions of others, they always try to understand the intentions underlying the actions. The neural mechanism of this understanding is referred to as the mirror neuron system (MNS). Different actions may correspond to different intentions, and the activation of the MNS in the human brain may also be slightly different. The present study distinguishes these differences according to functional brain imaging signals analyzed with machine learning. Brain signals were detected when the participants observed two types of actions: (1) grasping a cup for drinking, and (2) no meaningful contact. A synchronous measurement method for EEG and NIRS was adopted to increase the information contained in the brain signals. In order to obtain better classification accuracies, the method of functional brain networks was used. This method can be used to examine the relationship between brain regions. First, phase synchronization and Pearson correlation were used to calculate correlations for EEG channels and NIRS channels, respectively. Next, correlation matrices were converted into binary matrices, and the local properties of the networks were obtained. Finally, the feature vectors for the classifier were selected by analysis of their significance. In addition, EEG data and NIRS data were combined at the feature level and better classification results were obtained.
This paper presents the design of a modular time-frequency joint coding method for steady-state visual evoked potential, in which the stimulus consists of two stages: module flashing and individual flashing. In the first stage, targets are grouped into six-target modules. The modules flash at different frequencies and phases, whereas all targets within the module flash at the same frequency and phase. In the second stage, each target in the module flashes at different frequencies and phases. This coding method is compared with the control method in experiments. The results show that the proposed experimental paradigm performed better; the average accuracy and information transfer rate are higher than those of the control. Specifically, average accuracy is 89.6 ± 8.8%, which is 12.9% higher than that of the control experiment. These results prove the feasibility and superiority of the designed modular time-frequency joint coding method.