Recent advances in network technologies and Multi Agent Reinforcement Learning (MARL) have accelerated the development of Autonomous Underwater Vehicle (AUV) cluster networks, enabling intelligent applications such as target tracking and cooperative target encirclement. However, existing MARL models are typically designed for single-task scenarios, limiting their scalability in real-world multi-task environments. To ad dress this, we propose a lightweight MARL framework capable of handling multiple AUV tasks with reduced reliance on underwater sampling. Specifically, a unified state space representation is constructed to support task generalization, while a hybrid online offline MARL training paradigm is introduced by leveraging the logical reasoning and sample generation capabilities of Large Language Models (LLMs). This reduces the demand for real-time data collection. Furthermore, a supervised pretraining strategy is incorporated to improve convergence and learning stability. Based on these components, we develop the Large Language Model-driven Hybrid online-offline MARL algorithm towards Multi-Task scenarios (LLM-HMT), which supports intelligent deployment of multi-task AUV cluster systems with minimal state representation, reduced sample requirements, and limited training iterations. Extensive experiments demonstrate that LLM HMT outperforms mainstream MARL baselines in convergence speed, task success rate, and resource efficiency, highlighting its potential for practical underwater applications.

In recent years, the increasing number of industrial infrastructure security incidents around the world has drawn public attention to industrial control networks (ICNs) security issues. Fault diagnosis of industrial devices is an indispensable part of the security system in ICNs. The mainstream fault diagnosis models rely on long-term training and massive fault data, which results in the inability to update the model effectively and timely when the environment changes. Thus, some researchers focus on developing cross-domain industrial fault diagnosis methods. However, they usually presume that the samples of the target and source domains share the same fault mode sets, and existing prior knowledge concerning the label spaces of these two domains. These are difficult to satisfy in actual ICNs. To respond to these challenges, we develop atransferability-measured adversarial adaptation network(TAAN) to identify unknown classes without prior knowledge. It embeds the hybrid transferability estimation into an adversarial domain adaptive network to weigh the contribution of each sample. In this way, TAAN can properly classify samples in a public label space by selectively aligning source and target samples with high transferability. The experimental results obtained using two diagnosis datasets prove that the developed TAAN can achieve satisfactory diagnostic accuracy by effectively bridging the distribution discrepancy under various working conditions.

In information retrieval, relevance of documents with respect to queries is usually judged by humans, and used in evaluation and/or learning of ranking functions. Previous work has shown that certain level of noise in relevance judgments has little effect on evaluation, especially for comparison purposes. Recently learning to rank has become one of the major means to create ranking models in which the models are automatically learned from the data derived from a large number of relevance judgments. As far as we know, there was no previous work about quality of training data for learning to rank, and this paper tries to study the issue. Specifically, we address three problems. Firstly, we show that the quality of training data labeled by humans has critical impact on the performance of learning to rank algorithms. Secondly, we propose detecting relevance judgment errors using click-through data accumulated at a search engine. Two discriminative models, referred to as sequential dependency model and full dependency model, are proposed to make the detection. Both models consider the conditional dependency of relevance labels and thus are more powerful than the conditionally independent model previously proposed for other tasks. Finally, we verify that using training data in which the errors are detected and corrected by our method, we can improve the performance of learning to rank algorithms.

DNA Microarray technique has produced large amount of gene expression data. To analyze these data, many excellent machine learning techniques have been proposed in recent related work. In this paper, we try to perform the clustering of microarray data by combining the recently proposed Locality Preserving Projection (LPP) method with PCA, i.e. PCA-LPP. The comparison between PCA and PCA-LPP is performed based on two clustering algorithms, K-means and agglomerative hierarchical clustering. As we already known, clustering with the components extracted by PCA instead of the original variables does improve cluster quality. Moreover, our empirical study shows that by using LPP to perform further process the dimensions of components extracted by PCA can be further reduced and the quality of the clusters can be improved greatly meanwhile. Particularly, the first few components obtained by PCA-LPP capture more information of the cluster structure than those of PCA.

In the recent few years, web mining has become a hotspot of data mining with the development of Internet. Web pages classification is one of the essential techniques for web mining since classifying web pages of an interesting class is often the first step of mining the web. The high dimensional text vocabulary space is one of the main challenges of web pages. In this paper, we study the capabilities of Bayesian classifiers for web pages categorization. Several feature selection techniques, such as Chi Squared, Information Gain and Gain Ratio are used for selecting relevant words in web pages. Results on benchmark dataset show that the performances of Aggregating One-Dependence Estimators (AODE) and Hidden Naive Bayes (HNB) are both more competitive than other traditional methods.

Traditional linear discriminant analysis (LDA) is a popular dimensionality reduction method which preserve class separability. The method needs the labeled data to train. However in real worlds, the labeled training examples are very few but there are sufficient unlabeled data examples, so some former work (SDA) has made use the unlabeled training examples to do dimensionality reduction. But sometimes, there exits another kind of domain knowledge: the instances level constraints. In this paper, we consider the case when there are some useful instance level constraints. We propose a novel algorithm semi-supervised discriminant analyze with constraints (SDAC) which use three kinds of data: very few labeled data examples, sufficient unlabeled data examples and the instance level constraints. Our algorithm can be viewed as a constraint extension of traditional SDA algorithm. Experiments have been presented for semi-supervised classification tasks and have shown the effectiveness of our algorithm.

In this paper, we report our work on spam filtering with three novel Bayesian classification methods: aggregating one-dependence estimators (AODE), hidden Naive Bayes (HNB), locally weighted learning with Naive Bayes (LWNB). Other four traditional classifiers: Naive Bayes, k nearest neighbor (kNN), support vector machine (SVM), C4.5 are also performed for comparison. Four feature selection methods: gain ratio, information gain, symmetrical uncertainty and ReliefF, are used to select relevant words for spam filtering. Results of experiments on two corpora show the promising capabilities of Bayesian classifiers for spam filtering, especial for that of AODE.

In real world applications, there are great many of DNA expressed microarray data, many supervised classification algorithms such as decision tree, KNN and SVM in the machine learning field have been introduced for microarray data classification. However, in real worlds, the labeled examples, especially gene expression data examples are often very difficult and expensive to obtain. The traditional supervised methods can not work well when lack of training examples. So in this paper, we propose to use the semi-supervised learning algorithms which learning with both labeled and unlabeled data to do classification for microarray data. We perform experiments on four public microarray data sets and the results showed the semi-supervised method holds a much higher classification accuracy than the supervised methods and is much more stable when the labeled examples are very few.

In this paper, we propose a novel semi-supervised algorithm, which works under a two-view setting. Our algorithm, named Kernel Canonical Component Analysis Graph (KC-GRAPH), can effectively enhance the performance and the parameter stability of traditional graph-based semi-supervised algorithms by taking the advantage of two views using Kernel Canonical Component Analysis (KCCA). Experiments have been presented for semi-supervised classification tasks, and have shown that our KC-GRAPH algorithm stays a high classification accuracy and is much more stable than the former algorithms. We also noticed that our algorithm holds very good parameter stability.

Support vector machines have been extensively used in machine learning because of its efficiency and its theoretical background. This paper focuses on nu-transductive support vector machines for classification (nu-TSVC) and construct a new algorithm - Unconstrained nu-Transductive Support Vector Machines (Unu-TSVM). After researching on the special construction of primal problem in nu-TSVM, we transform it to an unconstrained problem and then smooth the derived problem in order to apply usual optimization methods. Numerical experiments prove its successful application in real life credit card dataset.

Recently increasing interests of applying or developing specialized machine learning techniques have attracted many researchers in the intrusion detection community. Existing research work show: the supervised algorithms deteriorates signifycantly if unknown attacks are present in the test data; the unsupervised algorithms exhibit no significant difference in performance between known and unknown attacks but their performances are not that satisfying. In this contribution, we propose two semi-supervised classification methods, Spectral Graph Transducer and Gaussian Fields Approach, to detect unknown attacks and one semi-supervised clustering method—MPCK-means to improve the performances of the traditional purely unsupervised clustering methods. Our empirical study shows that performances of semi-supervised classification methods are much better than those of supervised classifiers, and semi-supervised clustering method can improve purely unsupervised clustering methods markedly.

In real worlds applications, some former research papers have shown that manifold learning tries to discover the non-linear low-dimensional data manifold from a high-dimensional space. Many natural images and the face images are believed to be sampled from a manifold. In this paper, we try to investigate whether discovering such manifold can aid the semi-supervised learning algorithms. We propose a novel graph-based learning algorithm locality preserving graph-based semi-supervised method (LLGSM), which firstly use both labeled and unlabeled examples as unlabeled to discover the manifolds of the data samples and then use the projected labeled examples together with projected unlabeled ones to do classification. Experiments performed on some public image data sets have demonstrated the effectiveness of our algorithm.