Applied Soft Computing最新文献

Tabular data, organized like tables with rows and columns, is widely used. Existing models for tabular data synthesis often face limitations related to data size or complexity. In contrast, deep generative models, a part of deep learning, demonstrate proficiency in handling large and complex data sets. While these models have shown remarkable success in generating image and audio data, their application in tabular data synthesis is relatively new, lacking a comprehensive comparison with existing methods. To fill this gap, this study aims to systematically evaluate and compare the performance of deep generative models with these existing methods for tabular data synthesis, while also investigating the efficacy of post-processing techniques. We aim to identify strengths and limitations and provide insights for future research and practical applications. Our study showed that the Synthetic Minority Oversampling Technique (SMOTE) and its variants outperform deep generative models, especially for small datasets. However, we observed that an ensemble of deep generative models and post-generation processing performs better on large datasets than SMOTE alone. The results of our study indicate that deep generative models hold promise as a valuable tool for generating tabular data. Nonetheless, further research is warranted to enhance the performance of deep generative models and gain a comprehensive understanding of their limitations.

This paper proposes the probabilistic bipolar fermatean hesitant fuzzy set (PBFHFS) and its triangular version as generalizations of bipolar fuzzy set (BFS) for delineating randomness and imprecision in a single framework. To enhance pragmatic applicability, the devised set has been examined based on its fundamental operations, basic properties, and defuzzification mechanism. Offerings made to temples are considered sacred, so they cannot be disposed of in landfills. Waste from temples is usually disposed of in nearby lakes, ponds, and rivers. Thus, management of the temple floral waste (TFW) is essential to enhance the sustainability of the environment. In this regard, a green 4-dimensional transportation system under PBFHF settings is developed for the effective and sustainable managements of floral waste that is dumped from places of worship. For the most cost-effective and environmentally friendly results, this transportation system optimises a number of metrics, transportation expenditure (TE), carbon emission (CE), transportation time (TT), and labour cost (LC). Three sub-models are further developed from the proposed model under PBFHF conditions by incorporating the theory of carbon policy and risk assessments. The suggested transportation system is addressed by employing a fuzzy methodology, fuzzy programming (FP) as well as two non-fuzzy approaches, weighted sum technique (WST) and goal programming technique (GP). At last, there is a numerical computation along with comparison analysis, management insights, a conclusion with limitations, and a scope for further study.

Accurate, reliable, and efficient extraction of photovoltaic (PV) model parameters is an essential step towards PV system simulation, control, and optimization. Nevertheless, this problem is still facing great challenges because of its intrinsic nonlinear, multivariate, and multimodal properties. In this paper, a new variant of cuckoo search (CS), adaptive operator selection CS (AOSCS), is advanced for the PV model parameter extraction problems. AOSCS includes two major improvements: (1) an adaptive operator selection mechanism is developed to automatically assign the workloads of exploration and exploitation operators, and (2) the exploration and exploitation operators used in the original CS are modified to promote the exploration capability and reduce the blindness of search, respectively. The performance of AOSCS is firstly validated on CEC 2017 test suite and then it is utilized to solve the parameter extraction problems of five PV models. Moreover, further experiments on two commercial PV modules under distinct irradiance and temperature levels are also conducted to evaluate the practicality of the proposed algorithm. It is manifested that the results yielded by AOSCS are very competitive relative to other parameter extraction approaches. Accordingly, the proposed AOSCS is able to be served as an up-and-coming candidate algorithm for PV model parameter extraction problems.

Accurate prediction of short-term inbound passenger flow at high-speed railway stations is of great significance for the refined operation of stations, the formulation of emergency plans, and the provision of intelligent services. The arrival of passengers traveling on the same train at the same station shows a similar pattern, which is called the departure passenger arrival pattern (DPAP). The short-term inbound passenger flow at the station is composed of the short-term inbound passenger flow of all waiting trains within the same period. Inspired by this, this paper develops an ensemble prediction model based on the time series decomposition modeling strategy to introduce the DPAP to the short-term inbound passenger flow prediction at stations. Firstly, we propose a new framework for studying the DPAP to calculate the fitted station short-term inbound passenger flow, which is only affected by the DPAP. During this process, we find that 7 minutes is the optimal time granularity. Secondly, based on the singular spectrum analysis, we prove that the DPAP is the determining factor affecting the station short-term inbound passenger flow. Finally, we propose an ensemble prediction model that considers the DPAP to achieve short-term inbound passenger flow prediction at stations. The model consists of two parts: the deterministic and stochastic components prediction, where the former is predicted by the fitted station short-term inbound passenger flow, and the latter is achieved by the combination of historical stochastic components and weather type with the help of the Seq2Seq model based on time attention mechanism. Using real inbound passenger flow data, we compare the proposed model with 13 benchmark models and the results show that under different training and prediction steps, our model achieves optimal prediction performance, whether in all-day period and the busiest period of the station. Through further ablation experiments, it has been proven that the introduction of the DPAP effectively improves the prediction accuracy. Our model can provide scientific support for the intelligent operation of stations and the refined management of passenger flow.

The digital transformation process aims to align people, machines, and technologies in interactive systems, prompting businesses to adapt their models for competitiveness. Understanding and implementing digital transformation becomes crucial for organizations to integrate physical and virtual systems throughout their operations. The study presents a comprehensive model emphasizing intelligent technology, organization, and manufacturing as essential criteria. Integrating fuzzy cognitive maps and fuzzy DEMATEL methods, the study captures complex relationships and employs a dynamic analysis involving three firms in the defense industry. Results showcase the companies' diverse digital transformation levels, assessed through scenarios and predictive analyses. Additionally, a sensitivity analysis underscores the impact of adjustments to the top three criteria, aiding businesses in deciding their transformational starting points. This integrated method, underpinned by fuzzy cognitive maps and artificial intelligence, emerges as a decision support system and a roadmap for enterprises navigating digital transformation.

In the construction of information granule based neural networks for time series multi-step forecasting, the existing works tend to focus on consecutive-learning mode while rarely explore multi-learning mode. In fact, only under the multi-learning mode can diversified associations among data collected over time granules be well learned. Also, the existing works exhibit limited time interpretability. Here the problem centers around how to endow information granule based neural networks with a multi-learning mode to learn diversified associations simultaneously, and well-articulated trend semantics. To solve these problems, the first originality of this paper stems from a scale equalization method for multilinear-trend fuzzy information granules to track complex trend changes of data in a more accurate and explainable manner from both global and local views. Furthermore, an adaptive rather than empirical or traversal method, which is trend-driven in nature, is tailored for mining diversified associations. The resulting model can give forecasts in the form of granules as well as numerical values, being interpretable and accurate in the sense that: (a) its inputs and output are granules which come with well-defined trend semantics under a customary time concept, and (b) a clump of data is considered in a concise granule whilst roles of diversified associations are ware of during forecasting, making the model less prone to cumulative errors. Appealing experimental results corroborate the effectiveness of the proposed model.

Recently, Big data analytics approaches were combined with emerging machine learning techniques, which can provide more sophisticated insights into information-intensive activity. Compared to traditional shallow-architecture machine learning algorithms, Deep learning could excavate more potential information from the raw features. However, its powerful representational capacity relies on the support of enormous samples. The ensemble trees system performs superior on small sample problems due to better generalization capacity. To merge both benefits of deep learning and ensemble tree system, this paper developed a deep ensemble algorithm applied to multiple indicators prediction of pavement performance, including International Roughness Index and pavement 3-layer modulus. The deep ensemble algorithm is developed by merging a deep neural network (DNN) with the decision manifold property of the decision trees (TabNet) into a cascade ensemble system, combined with a sliding window algorithm to extract dependency information from raw data. During the training stage, the Bayesian Optimization Algorithm (BOA) is used to search for the optimal combination of sub-decision makers in the cascade ensemble. And equipped with GPU, it can speed up by 2.6–4.0 times. In the case study of pavement engineering, with sufficient training samples, it can achieve an average accuracy of 98.74 %, higher than DNN (97.49 %) and XGBoost (96.12 %) in predicting pavement indicators. With insufficient training samples, it can achieve an accuracy improvement of 12 % than XGBoost (75 %) and 24.5 % than DNN (62.5 %).

Biclusters hold significant importance in microarray analysis. Given the EA algorithm’s efficacy in tackling nonlinear problems, it has become a prevalent choice for evolutionary biclustering in microarray analysis. However, in conventional approaches, the objective of bicluster volume remains elusive, as it heavily relies on the yet-to-be-discovered real bicluster. This discrepancy introduces a novel research problem termed “unknown objective optimization” in our study. To address this issue, our paper introduces an innovative branching evolution strategy within a multi-objective framework. This strategy aims to resolve the challenge of unknown objectives. Throughout the biclustering search process, we meticulously observe the evolution of optimal bicluster individuals. Stability in both mean squared residue (MSR) and volume suggests a high likelihood of reaching an optimal solution, whether local or global. If a global optimal solution is attained at the end of the final evolution, our initial assumption is validated; otherwise, it necessitates an update. The proposed branching strategy is subsequently implemented to bifurcate the original evolution into two branches. One continues the original evolution with an unknown objective of bicluster volume, while the other pursues a new evolution with an estimated objective of bicluster volume. Our algorithm’s performance is assessed through comparisons with various traditional and evolutionary biclustering algorithms. The experimental results affirm its enhanced efficacy on both synthetic datasets and real gene datasets.

The application of artificial intelligence (AI) in the medical field has brought unprecedented opportunities and challenges for early diagnosis and precision treatment of cancer. As complex multi-omics data in the medical field tends to be multimodal, a single type of data cannot provide enough information to support accurate diagnosis. Vibrational spectroscopy consists of Raman spectroscopy and FTIR spectroscopy, both of which can reflect the structural information of molecules and are used to detect the vibration and rotational energy levels of material molecules. However, the application of multimodal tasks in fusing vibrational spectroscopy is not comprehensive. In response to the above problems, this paper focuses on interactive multimodal fusion strategies to process and mine vibrational spectral information. A Cross Branch Co-Attention Network (CBCAN) is proposed to solve the problem of insufficient spectral fusion, and a spectral branch network and a collaborative attention network are constructed for collaborative information fusion. Finally, feature-level fusion is combined to achieve better sequential decision-making effects. Extensive experiments were conducted on cancer datasets and thyroid dysfunction binary classification datasets, with the corresponding sample numbers of 192 and 379, respectively. The research results show that compared with traditional deep learning algorithms and the latest related multimodal medical fusion methods, the proposed CBCAN classification model achieved 96.88 % accuracy, 93.61 % precision, 91.52 % sensitivity, 98.03 % specificity, 91.73 % F1 score and 99.75 % AUC value, respectively, with the best classification effect, providing a new method for rapid and non-invasive identification of multiple selected cancers, which has important reference value for the early diagnosis of cancer patients and helps to assist clinical diagnosis.

Knowledge tracing (KT) is a crucial task in online education that traces students’ evolving cognition changes over time. However, it is a challenging task due to the heterogeneity of knowledge and incomplete cognition evolution sequences. This paper proposes KT-Deeper, a long-term Knowledge Tracing framework based on Dynamic reinforced heterogeneous graph contrastive networks, to predict students’ cognitive states on specific skills. Particularly, KT-Deeper initially employs temporal heterogeneous graphs to model the interconnections between different types of knowledge entities (e.g., students, exercises, and skills). Subsequently, KT-Deeper formalizes knowledge tracing as a dynamic link prediction task on the temporal heterogeneous graph sequence and proposes a reinforced graph generation approach to refine the incomplete graph sequence for supporting long-term knowledge tracing. KT-Deeper further presents a self-supervised heterogeneous graph embedding method to extract the structural features of knowledge evolution. Finally, KT-Deeper leverages recurrent neural networks to learn the temporal features of students’ cognitive evolution and predict whether a student will master a specific skill. Experimental results confirm that KT-Deeper exhibits superior performance compared to existing cutting-edge techniques, showcasing its promising accuracy and robustness in incomplete long-term knowledge tracing tasks.