In recent years, the Internet of Things has been applied in many fields with rapid development, such as software, sensors, and medical and healthcare. In the case of medical and healthcare, extensive research has focused on the development of brain-computer interface systems, particularly those utilizing steady-state visual-evoked potentials (SSVEPs). However, the conventional short-time Fourier transform (STFT) analysis is associated with the low-frequency resolution because of the length of the analysis window, resulting in sidelobe artifacts. In this paper, we utilized the non-harmonic analysis (NHA), which does not depend on the length of the analysis window, to analyze the continuous changes in and determine the classification accuracy of SSVEPs. Moreover, our experiments utilized the gray-scale images, allowing for the presentation of the stimulus as a sinusoidal pattern and reducing the effect of frequency distortion associated with the refresh rate of the liquid-crystal display. Our findings indicated that NHA resulted in exponential improvements in time-frequency resolution when compared with the STFT analysis. As the accuracy of NHA was high, our results suggest that this method is effective for examining SSVEPs and changes in brain waves during experiments conducted using liquid-crystal displays.

The frequency of the observed gravitational wave (GW) signal GW150914 changes slowly over a long time in the low frequency band and then increases rapidly from low to high frequency band, the frequency variance over time of such signals cannot be analyzed and visualized accurately using current analytical methods due to their fixed analytical window lengths or low time-frequency resolutions. To obtain high-accuracy GW representations, we propose the use of multiwindow nonharmonic analysis (NHA); in this method, NHA is performed using different window lengths for each subband after a band division procedure. A simulated signal was used to compare the results obtained using NHA and multiwindow NHA in this paper. The results distinctly show that the error of multiwindow NHA is 6.311 Hz which is lower than NHA. We also compared the results of multiwindow NHA with those of state-of-the-art methods by using actual GW150914 data and verified the analytical precision of these techniques. We found that the high timefrequency resolution of multiwindow NHA and its low sensitivity to analysis window variations enable it to analyze more accurately with 6.021 Hz errors in the extracted interval than the state-of-the-art methods even with the presence of significant noise. In addition, only multiwindow NHA can accurately represent GW signals at amplitudes down to the noise limit and achieve correct low-frequency range representation, due to its ability to assign the appropriate window length for each subband.

In some real-world optimization problems, the parameters of an objective function may be expressed as intervals, such as the benefit of a project and the driving speed of a robot. An optimization problem involving interval parameters and multiple conflicting objectives is termed as a multi-objective optimization problem with interval parameters (IMOP). Few studies have addressed IMOPs compared to deterministic multi-objective optimization at present. In addition, the uncertainty involved in the problems raises higher demands on the diversity and efficiency of an algorithm. Therefore, an adaptive reference vector-based interval multi-objective evolutionary algorithm in the framework of MOEA/D (IMOEA/D) was proposed in this paper. Firstly, an IMOP is decomposed into a number of sub-problems with interval parameters by setting an interval-valued reference point and constructing interval-valued scalar functions. Following that, an ensemble scheme for evaluating interval individuals is developed via the rank sum ratio method. Finally, reference vectors are adaptively adjusted based on the interval crowding distance to enhance the distribution of a population. The proposed IMOEA/D was tested on 20 benchmark IMOPs as well as a scheduling problem of underwater wireless sensor networks, and compared with four state-of-the-art interval MOEAs. The empirical results demonstrate its superior performance and strong competitiveness.

With the characteristics of simple structure and low cost, the dendritic neuron model (DNM) is used as a neuron model to solve complex problems such as nonlinear problems for achieving high-precision models. Although the DNM obtains higher accuracy and effectiveness than the middle layer of the multilayer perceptron in small-scale classification problems, there are no examples that apply it to large-scale classification problems. To achieve better performance for solving practical problems, an approximate Newton-type method-neural network with random weights for the comparison; and three learning algorithms including back-propagation (BP), biogeography-based optimization (BBO), and a competitive swarm optimizer (CSO) are used in the DNM in this experiment. Moreover, three classification problems are solved by using the above learning algorithms to verify their precision and effectiveness in large-scale classification problems. As a consequence, in the case of execution time, DNM + BP is the optimum; DNM + CSO is the best in terms of both accuracy stability and execution time; and considering the stability of comprehensive performance and the convergence rate, DNM + BBO is a wise choice.

In recent years, the dendritic neural model has been widely employed in various fields because of its simple structure and inexpensive cost. Traditional numerical optimization is ineffective for the parameter optimization problem of the dendritic neural model; it is easy to fall into local in the optimization process, resulting in poor performance of the model. This paper proposes an intelligent dendritic neural model firstly, which uses the intelligent optimization algorithm to optimize the model instead of the traditional dendritic neural model with a backpropagation algorithm. The experiment compares the performance of ten representative intelligent optimization algorithms in six classification datasets. The optimal combination of user-defined parameters for the model evaluates by using Taguchi’s method, systemically. The results show that the performance of an intelligent dendritic neural model is significantly better than a traditional dendritic neural model. The intelligent dendritic neural model has small classification errors and high accuracy, which provides an effective approach for the application of dendritic neural model in engineering classification problems. In addition, among ten intelligent optimization algorithms, an evolutionary algorithm called biogeographic optimization algorithm has excellent performance, and can quickly obtain high-quality solutions and excellent convergence speed.

In the field of image recognition, the all-MLP architecture (MLP-Mixer) shows superior performance. However, the current MLP-Mixer is solely based on fully connected layers. The nonlinear capability of fully connected layers is relatively weak, and their simple stacked structure has limitations under complex conditions. Therefore, inspired by the diversity of neurons in the human brain, we propose an innovative DMixNet, a dendritic multi-layered perceptron architecture. Rooted in the theory of dendritic neurons from neuroscience, we propose a dendritic neural unit (DNU) that enhances DMixNet with stronger biological interpretability and more robust nonlinear processing capabilities. The flexibility of dendritic structures allows the DNU to adjust its architecture to achieve different functionalities. Based on the DNU, we propose a novel channel fusion network \(\text {DNU}_\text {E}\) and a dendritic classifier \(\text {DNU}_\text {C}\). The \(\text {DNU}_\text {E}\) substitutes the traditional two fully connected layers as the channel mixer, constructing a dendritic mixer layer to enhance the fusion capability of channel information within the entire framework. Meanwhile, the \(\text {DNU}_\text {C}\) replaces the traditional linear classifier, effectively improving the model’s classification performance. Experimental results demonstrate that DMixNet achieves improvements of 2.13%, 4.79%, 4.71%, 23.14% on the CIFAR-10, CIFAR-100, Tiny-ImageNet and COIL-100 benchmark image recognition datasets, respectively, as well as a 14.78% enhancement on the medical image classification dataset PathMNIST, outperforming other state-of-the-art architectures. Code is available at https://github.com/KarilynXu/DMixNet.