Information Sciences最新文献

The process of training feedforward neural networks (FFNNs) can benefit from an automated process where the best heuristic to train the network is sought out automatically by means of a high-level probabilistic-based heuristic. This research introduces a novel population-based Bayesian hyper-heuristic (BHH) that is used to train feedforward neural networks (FFNNs). The performance of the BHH is compared to that of ten popular low-level heuristics, each with different search behaviours. The chosen heuristic pool consists of classic gradient-based heuristics as well as meta-heuristics (MHs). The empirical process is executed on fourteen datasets consisting of classification and regression problems with varying characteristics. The BHH is shown to be able to train FFNNs well and provide an automated method for finding the best heuristic to train the FFNNs at various stages of the training process.

The study demonstrates the application of OWA operators to binary and multiclass classification problems and seeks a way to improve classification accuracy using smoothing methods. OWA operators are used to aggregate class membership probabilities obtained from individual classifiers. Smoothing methods inspired by Newton-Cotes quadratures are applied before the aggregation step to improve the quality of the final results. Moreover, several sets of weights are used for OWA operators, including sets of weights based on the accuracy of individual classifiers. The experiments are conducted on 20 datasets, from which 7 are designed for binary classification and the rest are for multiclass classification. A comparison of the average accuracy for different sets of weights is shown. On the basis of experimental results, smoothing methods that significantly improve classification accuracy are identified.

As heavy-tailed distributions are ubiquitous in many real applications, robust regression has been extensively applied in machine learning and exhibits the superiority in deal with heavy-tailed distribution. Current existing robust methods for regression are based on independence assumption of features and ignore interaction between them, which may lead to poor generalization due to most datasets unsatisfying the assumption. Actually, interaction features formed by the composition of two or more features are particularly helpful to obtain the high-order information in many applications. In this paper, we propose a novel adaptive robust feature interaction selection model for regression with heavy-tailed distributions, termed Adaptive Robust Feature Interaction Selection (ARFIS). Firstly, we consider pairwise feature interaction by augmenting a feature vector with product of features for regression with heavy tailed distribution. Secondly, we propose feature interaction selection models based on quantile loss with different regularizers to learn parameters. The consistency of the proposed model ARFIS is theoretically proven, and an efficient algorithm is presented for solving proposed model. Finally, experimental results on simulation data, UCI datasets and a real-world dataset validate good accuracy, interpretability and robustness of our proposed models.

Developing an early warning model for company financial crises holds critical significance in robust risk management and ensuring the enduring stability of the capital market. Although the existing research has achieved rich results, the disadvantages of insufficient text information mining and poor model performance still exist. To alleviate the problem of insufficient text information mining, we collect related financial and annual report data from 820 listed companies in mainland China from 2018 to 2023 by using sophisticated web crawlers and advanced text sentiment analysis technologies and using missing value interpolation, standardization, and data balancing to build multi-source datasets of companies. Ranking the feature importance of multi-source data promotes understanding the formation of financial crises for companies. In the meantime, a novel Attention-based Fuzzy Neural Network (AFNN) was proposed to parse multi-source data to forecast financial crises among listed companies. Experimental results indicate that AFNN exhibits significantly improved performance compared to other advanced methods.

Brain age gap can be estimated from brain images, serving as a valuable biomarker for aging-associated diseases, using deep neural networks. Traditional brain age prediction methods tend to rely on unimodal data. Multimodal data can provide more comprehensive information and improve prediction accuracy. However, existing multimodal fusion methods often fall short in fully leveraging the correlations and complementarities between different modalities. This paper introduces a novel multimodal fuzzy feature fusion collaborative prediction algorithm for brain age estimation (MFCA). The proposed approach integrates multiple imaging modalities using a fuzzy fusion module and a multimodal collaborative convolutional module to effectively leverage inter-modal correlations and complementary information. Specifically, a convolutional neural network is used to extract feature from multimodal brain images, which are then combined into a global feature tensor via radial joins. The fuzzy fusion module employs fuzzy theory to fuse the correlation features of different modalities, while the multimodal collaborative convolutional module enhances complementary information through modality-specific convolutional layers. Age prediction is then performed by an age prediction module containing three linear regression modules. Additionally, an optimized sorting contrast loss is incorporated to improve the accuracy of age prediction. The proposed method was evaluated on the SRPBS multi-disorder MRI dataset, and the experimental results demonstrate that MFCA achieves a mean absolute error of 5.661 and a Pearson correlation coefficient of 0.947, outperforming several state-of-the-art methods.

A multivariable adaptive decoupling control scheme is proposed based on stochastic configuration networks with serial-parallel switching distribution (SPSCN). Firstly, a linear controller is designed by combining a PID controller, feedback decoupling, and one-step optimal control. Secondly, a nonlinear controller is presented to deal with higher-order nonlinear terms and unknown external perturbations. SPSCN is used to improve the prediction accuracy of unmodeled dynamics. It combines uniform and normal search strategies in a serial-parallel fashion, aiming at improving the node quality and reducing the model complexity. The approximation performance of the SPSCN algorithm is demonstrated by performing approximation experiments with two functions and four benchmark datasets. Compared with the generalized minimum variance control (GMVC) algorithm in controlling the process of cement raw material decomposition, our proposed control scheme outperforms.

Autonomous execution of tasks by unmanned aerial vehicles (UAVs) relies heavily on object detection. However, object detection in most images presents challenges such as complex backgrounds, small targets, and obstructions. Additionally, the limited computing speed and memory of the UAV processor affects the accuracy of conventional object detection algorithms. This paper proposes LUD-You Only Look Once (YOLO), a small and lightweight object detection algorithm for UAVs based on YOLOv8. The proposed algorithm introduces a new multiscale feature fusion mode that solves the degradation in feature propagation and interaction through the introduction of upsampling in the feature pyramid network and the progressive feature pyramid network. The application of the dynamic sparse attention mechanism in the Cf2 module achieves flexible computing allocation and content awareness. Furthermore, the proposed model is optimized to be sparse and lightweight, making it possible to deploy on UAV edge devices. Finally, the effectiveness and superiority of LUD-YOLO were verified on the VisDrone2019 and UAVDT datasets. The results of ablation and comparison experiments show that compared with the original algorithm, LUDY-N and LUDY-S have shown excellent performance in various evaluation indexes, indicating that the proposed improvement strategies make the model have better robustness and generalization. Moreover, compared with multiple other popular competitors, the proposed improvement strategies enable LUD-YOLO to have the best overall performance, providing an effective solution for UAVs object detection while balancing model size and detection accuracy.

The preference information of decision-makers (DMs) often cannot be explicitly expressed in complex decision-making environments. Therefore, to address decision-making problems with uncertain preference information, this paper proposes a method based on the prospect SMAA-2 model and extended cross-entropy in interval number environments. We first construct the prospect SMAA-2 model to simulate preference information. This model incorporates central risk-averse and risk-seeking factors, significantly enhancing the ability to identify alternatives. When DMs’ preference information is unknown or partially known, these factors can determine the appropriate level of risk-averse or risk-seeking for alternatives. Next, we devise an extended cross-entropy algorithm based on the continuous ordered weighted harmonic (C-OWH) averaging operator to handle interval numbers. Subsequently, a comprehensive algorithm is designed to derive the weights of DMs, taking into account the relationships among individuals as well as between individuals and the group. Furthermore, we construct the framework for the proposed method. Finally, the applicability of the developed method can be validated by an illustrative example. Comparative analysis is used to verify the rationality and superiority of this method.

In existing trust-aware collaborative filtering algorithms, each trust relationship between two users is usually represented by a real number, but such a number is neither sufficient to reflect the quantity of the trust relationship existing in the user’s mind nor easy to be given. This leads to the inaccuracy of the trust relationship and poor final recommendations. To solve this problem, we propose an approach to deduce interval-valued trust relationships from the given real-valued trust relationships, which enables the new trust relationships to optimally reflect the true trust relationships existing in users’ minds. The coming problem we face is how to fuse the interval-valued trust relationships and the real-valued ratings. Though most existing trust-aware collaborative filtering algorithms use matrix factorization to fuse the real-valued data, they are not capable of fusing interval-valued trust relationships and real-valued ratings. The reason is that the arithmetic operations on intervals and arithmetic operations on real numbers are different. Therefore, we proposed a novel interval-valued matrix factorization approach. After that, an interval-valued matrix factorization based trust-aware collaborative filtering (IMF_TCF) algorithm is designed. The experiments carried out with open datasets indicate that IMF_TCF achieves the best recommendation performance compared with the state-of-the-art algorithms.

To benefit from the modeling capacity of deep models in system identification without worrying about inference time, this study presents a novel training strategy that uses deep models only during the training stage. For this purpose, two separate models with different structures and goals are employed. The first one is a deep generative model aiming at modeling the distribution of system output(s), called the teacher model, and the second one is a shallow basis function model, named the student model, fed by system input(s) to predict the system output(s). That means these isolated paths must reach the same ultimate target. As deep models show a great performance in modeling highly nonlinear systems, aligning the representation space learned by these two models makes the student model inherit the teacher model’s approximation power. The proposed objective function consists of the objective of each student and teacher model, adding up with a distance penalty between the learned latent representations. The simulation results on three nonlinear benchmarks show a comparative performance with examined deep architectures applied on the same benchmarks. Algorithmic transparency and structure efficiency are also achieved as byproducts.