Recent autonomous driving systems heavily rely on 3D point cloud data collected from multiple sensors for environmental awareness and decision-making. However, it is unrealistic to expect the autonomous driving system to recognize all road environments and handle every traffic situation. Models for autonomous driving need to be updated in real time in order for the system to adapt to more situations. This is where online continual learning becomes crucial. Online continual learning is an important method in the field of autonomous driving, as it enables models to update their parameters with streaming input data for adapting to new environments and conditions. Online continual learning in the field of autonomous driving faces several challenges: inefficient data fusion, catastrophic forgetting, insufficient computational resources, violation of road privacy and categories imbalance. To tackle these challenges, we propose an Analytic Online Continual Learning method for 3D Point Cloud Classification (3D-AOCL). This approach utilizes Analytic Learning to update parameters and integrates a feature fusion module along with a category balancer to address the above issues. It is capable of fusing data in feature level, balancing samples across various categories and updating parameters by calculating the analytical solution. We have validated our method on the vehicle side, the infrastructure side, and vehicle-infrastructure cooperative data on the V2X-Seq dataset. The experimental results demonstrate that our model effectively addresses key issues in online continual learning for autonomous driving systems, outperforming other models by approximately 4.00% to 6.00% in AMCA scores while only keeping 0.75% trainable parameters.

Syndrome classification plays a key role in the clinical diagnosis and treatment with traditional Chinese medicine (TCM), aiming to identify the disease type. Given symptoms of a patient, existing approaches for syndrome classification are generally limited to modeling the interaction between symptoms and syndromes while ignoring the induction of state elements. To alleviate this issue, a state-element-aware hypergraph convolutional network (SEHGCN) is proposed to incorporate state elements into syndrome classification and discover high-order semantic relationships among TCM entities through hypergraph convolutional network (HGCN). Specifically, state elements are initially induced from symptoms by an extraction network, then symptoms and state elements are embedded via convolution on patient hypergraph to obtain the latent representation of patients. Finally, syndromes are classified by a multilayer perceptron (MLP). Extensive experiments on two TCM datasets show that the syndrome classification results with this proposed method are significantly improved over other competing methods.

Jailbreaking attacks can enable Large Language Models (LLMs) to bypass the safeguard and generate harmful content. Existing jailbreaking defense methods have failed to address the fundamental issue that harmful knowledge resides within the model, leading to potential jailbreak risks for LLMs. In this paper, we propose a novel defense method called Eraser, which mainly includes three goals: unlearning harmful knowledge, retaining general knowledge, and maintaining safety alignment. The intuition is that if an LLM forgets the specific knowledge required to answer a harmful question, it will no longer have the ability to answer harmful questions. The training of Erase does not actually require the model's own harmful knowledge, and it can benefit from unlearning general answers related to harmful queries, which means it does not need assistance from the red team. The experimental results show that Eraser can significantly reduce the jailbreaking success rate for various attacks without compromising the general capabilities of the model.

Event argument extraction (EAE) is an important task in information extraction. Traditional supervised methods are not easy to generalize to new event types. Most zero-shot EAE methods learned event-specific argument representation or event-specific role representation, which makes it difficult to generalize to new event types. We propose Distar, a zero-shot EAE method that disentangles the trigger from the argument and the role by modeling them using a knowledge graph embedding algorithm, i.e., TransE. We treat the argument as the head entity, the role as the relation, and the trigger as the tail entity. To enhance the generalization capacity, we construct two classifiers to relate the new role to the existing ontology. The experimental results show that our method can outperform the baseline by 8 https://github.com/ZeroNLP/Distar for reproduction purposes.

Introduction: Traditional Chinese Medicine (TCM) diagnosis is a reasoning process through expert knowledge, in which syndrome classification is a key step for prescription recommendation and the treatment of patients. Doctors generally differentiate syndrome types according to patients' symptoms and state elements. This paper proposes a syndrome classification method based on graph convolutional network with residual structure, to exploit the potential relationship between symptoms and state elements. Methods: We constructed a graph convolutional network by combining symptoms and state elements for syndrome classification, called Symptoms-State elements Graph Convolutional Network (SSGCN), embedding the inherent logic of TCM diagnosis and treatment with a prescription graph. This graph architecture wherein contained the relationship between symptoms and state elements, and a multi-layer perceptron (MLP) was trained to classify different syndromes.Results: Experiments were conducted on two self-built datasets according to two classic TCM books, i.e., Theories on Febrile Diseases and Traditional Chinese Medicine Prescription Dictionary. Accuracy, precision, recall and F1 score were adopted to evaluate the syndrome classificaiton results. Our proposed SSGCN method achieved accuracy of 75.59%, 69.63%, precision of 69.10%, 76.33%, recall of 75.63%, 66.67% and F1-score of 71.26%, 65.84% in the above two datasets, respectively. The proposed method for syndrome classification outperformed several popular methods including support vector machine, random forest, extreme gradient boosting and convolutional neural network.Conclusions: By constructing a prescription graph in which symptoms are used as nodes and state elements are taken into account for edges, graph convolution is implemted to capture the relationship of symptoms and state elements. This model improves the performance of syndrome classification and can be further extened for some other related applications in TCM.

An accurate tourism demand forecasting model is crucial for tourism decision-makers. In recent years, several deep learning-based models have been developed to predict tourist arrivals via search intensity indices. However, few methods consider the lag effect in the long-term time range and the interaction between different search intensity indices factors. To alleviate the above limitations, we propose a graph-guided tourism demand forecasting network, which can model the lag effect of historical variables on future variables. Specifically, each variable is individually encoded via a convolutional neural network in the time dimension. Then, lag effects are modeled dynamically in a bipartite graph, and mined via graph aggregation. Finally, a fully-connected network is designed for regression prediction. Experimental results on two public datasets demonstrate the superiority of the proposed method in both one-step and multi-step prediction compared with existing methods.

Objective: To establish a state identification algorithm model using convolutional neural network model (CNN). Methods: Based on the fundamental principles of traditional Chinese medicine state identification, the study starts with state characterizations from medical cases and establishes a state characterization database. A CNN-based state element induction model (CNN-SEI) is constructed to connect the state characterizations to state elements and to identify states. Results: The model exhibits significant advantages in accuracy and recall rates, and the results of state identification are highly consistent with clinical diagnosis. Conclusion: This algorithm can improve the accuracy of state identification and provide guidance for clinical syndrome differentiation.

Intelligent fault diagnosis has been widely used in the industry and plays a crucial role in the health management of machinery. In recent years, unsupervised domain adaptation (UDA) has been applied to fault diagnosis, showing excellent performance under variable working conditions. However, most existing UDA-based methods do not consider the temporal relations in the fault signal, resulting in sub-optimal performance. In this paper, we proposed a domain adversarial interaction network (DAIN) to solve this problem. By downsampling sub-sequences of fault signals and interacting with their features, DAIN can obtain feature representations containing the temporal relations. In addition, domain adversarial learning and maximum mean discrepancy (MMD) are applied to DAIN to align the domain discrepancy and distribution discrepancy of source and target domains. We conducted extensive experiments on the public Paderborn University (PU) dataset, and the results demonstrate that the proposed method can achieve higher cross-domain fault diagnosis accuracy than the existing methods.

Improving the prediction accuracy of water levels is of great significance to water resources scheduling, flood control, and disaster reduction. Deep learning-based recurrent neural networks variants have proved to be a powerful algorithmic model for water level prediction. However, few people use generation adversarial network to mine water level time series information. In this work, we propose a simple and effective method (termed WLP-ForGAN) for watershed multi-point Water Level Prediction based on Conditional Generative Adversarial Networks. Through the adversarial training of the generator and discriminator, the predicted data is constantly close to the real data, to realize the prediction of time series, and improve the utilization of limited data information of the model. The attention module is embedded in the WLP-ForGAN network structure of the generator and discriminator, so that the WLP-ForGAN model can autonomously assign the attention weight value to the input series vectors according to the hydrological information in the input features. Extensive experiments on real datasets show that the proposed WLP-ForGAN consistently outperforms commonly used models, which proves the effectiveness of the proposed method.

Unsupervised domain adaptation-based fault diagnosis methods have been extensively studied due to their powerful knowledge transferability under different working conditions. Despite their encouraging performance, most of them cannot sufficiently account for the temporal dimension of the vibration signal, resulting in incomplete feature information used in the domain alignment procedure. To alleviate the limitation, we present a self-supervised domain adaptation fault diagnosis network (SDAFDN), which considers two temporal dependencies to improve the transferability of the learned representations. Specifically, we first design a down-sampling and interaction network that considers the temporal dependency among subsequences with low temporal resolution in feature space. Then, we combine domain adversarial learning with feature mapping to achieve domain alignment. Finally, we introduced a self-supervised learning module, which considers the temporal dependency between the past and future temporal segments via classification tasks. Extensive experiments on public Paderborn University and PHM data sets demonstrate the superiority of the proposed SDAFDN and the effectiveness of considering temporal dependencies in domain alignment.

Syndrome classification is an important step in Traditional Chinese Medicine (TCM) for diagnosis and treatment. In this paper, we propose a multi-graph attention network (MGAT) based method to simulate TCM doctors to infer the syndromes. Specifically, the complex relationships between symptoms and state elements are aggregated using graph attention networks (GAT) and syndromes are classified by a multilayer perceptron (MLP). To verify the effectiveness of the model, extensive experiments are conducted on the Treatise on Febrile Diseases dataset. The experimental results show that the proposed method outperforms several typical methods in terms of accuracy, precision, recall, and F1-score. The MGAT model has high accuracy in syndrome classification and has practical application value.

Ethnopharmacological relevance: The recommendation of herbal prescriptions is a focus of research in traditional Chinese medicine (TCM). Artificial intelligence (AI) algorithms can generate prescriptions by analysing symptom data. Current models mainly focus on the binary relationships between a group of symptoms and a group of TCM herbs. A smaller number of existing models focus on the ternary relationships between TCM symptoms, syndrome-types and herbs. However, the process of TCM diagnosis (symptom analysis) and treatment (prescription) is, in essence, a "multi-ary" (n-ary) relationship. Present models fall short of considering the n-ary relationships between symptoms, state-elements, syndrome-types and herbs. Therefore, there is room for improvement in TCM herbal prescription recommendation models. Purpose: To portray the n-ary relationship, this study proposes a prescription recommendation model based on a multigraph convolutional network (MGCN). It introduces two essential components of the TCM diagnosis process: state-elements and syndrome-types. Methods: The MGCN consists of two modules: a TCM feature-aggregation module and a herbal medicine prediction module. The TCM feature-aggregation module simulates the n-ary relationships between symptoms and prescriptions by constructing a symptom-'state element'-symptom graph (S-e) and a symptom-'syndrome-type'-symptom graph (T-s). The herbal medicine prediction module inputs state-elements, syndrome-types and symptom data and uses a multilayer perceptron (MLP) to predict a corresponding herbal prescription. To verify the effectiveness of the proposed model, numerous quantitative and qualitative experiments were conducted on the Treatise on Febrile Diseases dataset. Results: In the experiments, the MGCN outperformed three other algorithms used for comparison. In addition, the experimental data shows that, of these three algorithms, the SVM performed best. The MGCN was 4.51%, 6.45% and 5.31% higher in Precision@5, Recall@5 and F1-score@5, respectively, than the SVM. We set the K-value to 5 and conducted two qualitative experiments. In the first case, all five herbs in the label were correctly predicted by the MGCN. In the second case, four of the five herbs were correctly predicted. Conclusions: Compared with existing AI algorithms, the MGCN significantly improved the accuracy of TCM herbal prescription recommendations. In addition, the MGCN provides a more accurate TCM prescription herbal recommendation scheme, giving it great practical application value.

Objective: To develop a neural network model that recommends traditional Chinese medicine (TCM) herbal prescriptions. Methods: We constructed a new dataset of diagnosis and treatment knowledge from the Treatise on Febrile Diseases. Based on TCM's logical principles of “syndrome differentiation” and “state recognition”, a back-propagation neural network model is proposed that simulates clinical diagnosis and treatment. Results: The proposed model is a four-layer BP neural network. Experiments on the constructed dataset show that the proposed method achieved the best precision, recall, and F1-scores. Conclusion: The proposed method provides much more accurate herbal prescription recommendations than logistic regression.