Artificial Intelligence Review最新文献

Anomaly detection in streaming data is an important task for many real-world applications, such as network security, fraud detection, and system monitoring. However, streaming data often exhibit concept drift, which means that the data distribution changes over time. This poses a significant challenge for many anomaly detection algorithms, as they need to adapt to the evolving data to maintain high detection accuracy. Existing streaming anomaly detection algorithms lack a unified evaluation framework that validly assesses their performance and robustness under different types of concept drifts and anomalies. In this paper, we conduct a systematic technical review of the state-of-the-art methods for anomaly detection in streaming data. We propose a new data generator, called SCAR (Streaming data generator with Customizable Anomalies and concept dRifts), that can synthesize streaming data based on synthetic and real-world datasets from different domains. Furthermore, we adapt four static anomaly detection models to the streaming setting using a generic reconstruction strategy as baselines, and then compare them systematically with 9 existing streaming anomaly detection algorithms on 76 synthesized datasets that have various types of anomalies and concept drifts. The challenges and future research directions for anomaly detection in streaming data are also presented.

Sentiment analysis in natural language processing (NLP) is used to understand the polarity of human emotions (e.g., positive and negative) and preferences (e.g., price and quality). Reinforcement learning (RL) enables a decision maker (or agent) to observe the operating environment (or the current state) and select the optimal action to receive feedback signals (or reward) from the operating environment. Deep reinforcement learning (DRL) extends RL with deep neural networks (i.e., main and target networks) to capture the state information of inputs and address the curse of dimensionality issue of RL. In sentiment analysis, RL and DRL reduce the need for a large labeled dataset and linguistic resources, increasing scalability and preserving the context and order of logical partitions. Through enhancement, the RL and DRL algorithms identify negations, enhance the quality of the generated responses, predict the logical partitions, remove the irrelevant aspects, and ultimately capture the correct sentiment polarity. This paper presents a review of RL and DRL models and algorithms with their objectives, applications, datasets, performance, and open issues in sentiment analysis.

This empirical investigation examines the complex dynamics between Artificial Intelligence (AI), Potential Development (PD), Training Initiatives (TI), and High-Performance Work Systems (HPWS) within manufacturing firms to gain valuable insights into how AI technologies influence high-performance work systems through employee development and training. Using a purposive sampling technique, around two hundred employees from twenty-four manufacturing firms in the textile, automotive, steel, and pharmaceutical sectors participated in the self-administered survey. The empirical analysis of the data sets was conducted using the PLS-SEM approach. This result demonstrated positive associations between AI, PD, and HPWS, emphasizing the key role of AI in supporting employee development and improving high-performance work systems. Furthermore, training’s amplification effect on the relation between artificial intelligence and professional development highlighted the significance of employees’ upskilling for AI integration. Conversely, the mediating role of PD between AI adoption and HPWS effectiveness highlighted the significant role of employee professional development in achieving HPWS through AI integration within the systems. The study offered insight into the mediation of PD between AI and HPWS effectiveness, emphasizing its centrality in translating AI-driven advances into tangible organizational outcomes. The study findings have significant ramifications for both theory and practice. Theoretically, this research adds to an evolving dialogue surrounding AI’s effects on HR practices and organizational outcomes; practically speaking, organizations can utilize this research’s insights in strategically integrating AI technologies, designing tailored training programs for their employees, and creating an environment conducive to ongoing employee development.

Rough set theory is an important approach to deal with uncertainty in data mining. However, Pawlak’s classical rough set has low fault-tolerance on concept approximation based on knowledge granules, which may influence the classification accuracy in practical application. To address this problem, the present paper proposes a novel sequential rough-set model that is proved to be a conservative extension of Pawlak’s classical rough set. As a result, it effectively improves the fault-tolerance ability, classification accuracy and concept approximation accuracy of the latter without any additional assumption. Based on the properties and theoretical analysis of the proposed model, an algorithm is presented to automatically determine the sequential thresholds and compute the three regions for the given concept. Experiments on real data verify the validity of the algorithm, and also show the stable improvement on the two types of accuracy.

Emotion Recognition in Conversations (ERC) is a key step towards successful human–machine interaction. While the field has seen tremendous advancement in the last few years, new applications and implementation scenarios present novel challenges and opportunities. These range from leveraging the conversational context, speaker, and emotion dynamics modelling, to interpreting common sense expressions, informal language, and sarcasm, addressing challenges of real-time ERC, recognizing emotion causes, different taxonomies across datasets, multilingual ERC, and interpretability. This survey starts by introducing ERC, elaborating on the challenges and opportunities of this task. It proceeds with a description of the emotion taxonomies and a variety of ERC benchmark datasets employing such taxonomies. This is followed by descriptions comparing the most prominent works in ERC with explanations of the neural architectures employed. Then, it provides advisable ERC practices towards better frameworks, elaborating on methods to deal with subjectivity in annotations and modelling and methods to deal with the typically unbalanced ERC datasets. Finally, it presents systematic review tables comparing several works regarding the methods used and their performance. Benchmarking these works highlights resorting to pre-trained Transformer Language Models to extract utterance representations, using Gated and Graph Neural Networks to model the interactions between these utterances, and leveraging Generative Large Language Models to tackle ERC within a generative framework. This survey emphasizes the advantage of leveraging techniques to address unbalanced data, the exploration of mixed emotions, and the benefits of incorporating annotation subjectivity in the learning phase.

The industrialization, high-density, and greener aquaculture requires a more precise and intelligent aquaculture management. Phenotypic and behavioral information of fish, which can reflect fish growth and welfare status, play a crucial role in aquaculture management. Stereo vision technology, which simulates parallax perception of the human eye, can obtain the three-dimensional phenotypic characteristics and movement trajectories of fish through different types of sensors. It can overcome the limitations in dealing with fish deformation, frequent occlusions and understanding three-dimension scenes compared to the traditional two-dimensional computer vision techniques. With the deep learning development and application in aquaculture, stereo vision has become a super computer vision technology that can provide more precise and interpretable information for intelligent aquaculture management, such as size estimation, counting and behavioral analysis of fish. Hence, it is very beneficial for researchers, managers, and entrepreneurs to possess a thorough comprehension about the fast-developing stereo vision technology for modern aquaculture. This study provides a critical review of relevant topics, including the four-layer application structure of stereo vision technology in aquaculture, various deep learning-based technologies used, and specific application scenarios. The review contributes to research development by identifying the current challenges and provide valuable suggestions for future research directions. This review can serve as a useful resource for developing future studies and applications of stereo vision technology in smart aquaculture, focusing on phenotype feature extraction and behavioral analysis of fish.

The field of natural language processing has experienced significant advances in recent years, but these advances have not yet resulted in improved analytics for instructors on MOOC platforms. Valuable information, such as suggestions, is generated in the comment forums of these courses, but due to their volume, manual processing is often impractical. This study examines the feasibility of fine-tuning and effectively utilizing state-of-the-art deep learning models to identify comments that contain suggestions in MOOC forums. The main challenges encountered are the lack of labeled datasets from the MOOC context for fine-tuning classification models and the soaring computational cost of this training. For this study, we manually collected and labeled 2228 comments in Spanish and English from 5 MOOCs and scraped 1.4 million MOOC reviews from 3 platforms. We fine-tuned and evaluated 4 pretrained models based on the transformer architecture and 3 traditional machine learning models to compare their effectiveness in the suggestion mining task in this domain. Transformer-based models proved to be highly effective in this task/domain combination, achieving performance levels that matched or exceeded those deemed appropriate in other contexts and were significantly greater than those achieved by traditional models. Domain adaptation led to improved linguistic understanding of the target domain; however, in this project, this approach did not translate into an observable improvement in suggestion mining. The automated identification of comments that can be labeled as suggestions can result in considerable time savings for instructors, especially considering that less than a quarter of the analyzed comments contain suggestions.

The optimization challenge known as the optimal reactive power dispatch (ORPD) problem is of utmost importance in the electric power system owing to its substantial impact on stability, cost-effectiveness, and security. Several metaheuristic algorithms have been developed to address this challenge, but they all suffer from either being stuck in local minima, having an insufficiently fast convergence rate, or having a prohibitively high computational cost. Therefore, in this study, the performance of four recently published metaheuristic algorithms, namely the mantis search algorithm (MSA), spider wasp optimizer (SWO), nutcracker optimization algorithm (NOA), and artificial gorilla optimizer (GTO), is assessed to solve this problem with the purpose of minimizing power losses and voltage deviation. These algorithms were chosen due to the robustness of their local optimality avoidance and convergence speed acceleration mechanisms. In addition, a modified variant of NOA, known as MNOA, is herein proposed to further improve its performance. This modified variant does not combine the information of the newly generated solution with the current solution to avoid falling into local minima and accelerate the convergence speed. However, MNOA still needs further improvement to strengthen its performance for large-scale problems, so it is integrated with a newly proposed improvement mechanism to promote its exploration and exploitation operators; this hybrid variant was called HNOA. These proposed algorithms are used to estimate potential solutions to the ORPD problem in small-scale, medium-scale, and large-scale systems and are being tested and validated on the IEEE 14-bus, IEEE 39-bus, IEEE 57-bus, IEEE 118-bus, and IEEE 300-bus electrical power systems. In comparison to eight rival optimizers, HNOA is superior for large-scale systems (IEEE 118-bus and 300-bus systems) at optimizing power losses and voltage deviation; MNOA performs better for medium-scale systems (IEEE 57-bus); and MSA excels for small-scale systems (IEEE 14-bus and 39-bus systems).

In the evolving field of decision-making, the continuous advancement of technologies and methodologies drives the pursuit of more reliable tools. Decision support systems (DSS) provide information to make informed choices and multi-criteria decision analysis (MCDA) methods are an important component of defining decision models. Despite their usefulness, there are still challenges in making robust decisions in dynamic environments due to the varying performance of different MCDA methods. It creates space for the development of techniques to aggregate conflicting results. This paper introduces a fuzzy ranking approach for aggregating results from multi-criteria assessments, specifically addressing the limitations of current result aggregation techniques. Unlike conventional methods, the proposed approach represents rankings as fuzzy sets, providing detailed insights into the robustness of decision problems. The study uses green supplier selection as a case study, examining the performance of the introduced approach and the robustness of its recommendations within the sustainability field. This study offers a new methodology for aggregating results from multiple evaluation scenarios, thereby enhancing decision-maker awareness and robustness. Through comparative analysis with traditional compromise solution methods, this paper highlights the limitations of current approaches and indicates the advantages of adopting fuzzy ranking aggregation. This study significantly advances the field of decision-making by enhancing the understanding of the stability of decision outcomes.

The sand cat swarm optimization algorithm (SCSO) is a metaheuristic algorithm proposed by Amir Seyyedabbasi et al. SCSO algorithm mimics the predatory behavior of sand cats, which gives the algorithm a strong optimized performance. However, as the number of iterations of the algorithm increases, the moving efficiency of the sand cat decreases, resulting in the decline of search ability. The convergence speed of the algorithm gradually decreases, and it is easy to fall into local optimum, and it is difficult to find a better solution. In order to improve the search and movement efficiency of the sand cat, and enhance the global optimization ability and convergence performance of the algorithm, an improved sand cat Swarm Optimization (ISCSO) algorithm was proposed. In ISCSO algorithm, we propose a low-frequency noise search strategy and a spiral contraction walking strategy according to the habit of sand cat, and add random opposition-based learning and restart strategy. The frequency factor was used to control the search direction of the sand cat, and the spiral contraction hunting was carried out, which effectively improved the randomness of the population, expanded the search range of the algorithm, enhanced the moving efficiency of the sand cat, and accelerated the convergence speed of the algorithm. We use 23 standard benchmark functions and IEEE CEC2014 benchmark functions to compare ISCSO with 10 algorithms, and prove the effectiveness of the improved strategy. Finally, ISCSO was evaluated using five constrained engineering design problems. In the results of these problems, using ISCSO has 3.08%, 0.23%, 0.37%, 22.34%, 1.38% improvement compared with the original algorithm respectively, which proves the effectiveness of the improved strategy in practical application problems. The source code website for ISCSO is https://github.com/Ruiruiz30/ISCSO-s-code.