Nowadays, data parallelism has been widely applied to train large datasets on distributed deep learning clusters, but it has suffered from costly global parameter updates at batch barriers. Performance imbalance among...
详细信息
Event datasets in the financial domain are often constructed based on actual application scenarios, and their event types are weakly reusable due to scenario constraints;at the same time, the massive and diverse new f...
详细信息
In-context learning is a promising approach for offline reinforcement learning (RL) to handle online tasks, which can be achieved by providing task prompts. Recent works demonstrated that in-context RL could emerge wi...
In-context learning is a promising approach for offline reinforcement learning (RL) to handle online tasks, which can be achieved by providing task prompts. Recent works demonstrated that in-context RL could emerge with self-improvement in a trial-and-error manner when treating RL tasks as an across-episodic sequential prediction problem. Despite the self-improvement not requiring gradient updates, current works still suffer from high computational costs when the across-episodic sequence increases with task horizons. To this end, we propose an In-context Decision Transformer (IDT) to achieve self-improvement in a high-level trial-and-error manner. Specifically, IDT is inspired by the efficient hierarchical structure of human decision-making and thus reconstructs the sequence to consist of high-level decisions instead of low-level actions that interact with environments. As one high-level decision can guide multi-step low-level actions, IDT naturally avoids excessively long sequences and solves online tasks more efficiently. Experimental results show that IDT achieves state-of-the-art in long-horizon tasks over current in-context RL methods. In particular, the online evaluation time of our IDT is 36× times faster than baselines in the D4RL benchmark and 27× times faster in the Grid World benchmark.
Image segmentation is a crucial task in the field of computer vision. Markov random fields (MRF) based image segmentation method can effectively capture intricate relationships among pixels. However, MRF typically req...
详细信息
Graph neural network(GNN) is a promising method to analyze graphs. Most existing GNNs adopt the class-balanced assumption, which cannot deal with class-imbalanced graphs well. The oversampling technique is effective i...
详细信息
Graph neural network(GNN) is a promising method to analyze graphs. Most existing GNNs adopt the class-balanced assumption, which cannot deal with class-imbalanced graphs well. The oversampling technique is effective in alleviating class-imbalanced problems. However, most graph oversampling methods generate synthetic minority nodes and their edges after applying GNNs. They ignore the problem that the representations of the original and synthetic minority nodes are dominated by majority nodes caused by aggregating neighbor information through GNN before oversampling. In this paper, we propose a novel graph oversampling framework, termed distribution alignment-based oversampling for node classification in classimbalanced graphs(named Graph-DAO). Our framework generates synthetic minority nodes before GNN to avoid the dominance of majority nodes caused by message passing in GNNs. Additionally, we introduce a distribution alignment method based on the sum-product network to learn more information about minority nodes. To our best knowledge, it is the first to use the sum-product network to solve the class-imbalanced problem in node classification. A large number of experiments on four real datasets show that our method achieves the optimal results on the node classification task for class-imbalanced graphs.
Semi-supervised multi-label learning (SSMLL) refers to inducing classifiers using a small number of samples with multiple labels and many unlabeled samples. The prevalent solution of SSMLL involves forming pseudo-labe...
Transformer-based models are dominating the field of natural language processing and are becoming increasingly popular in the field of computer vision. However, the black box characteristics of transformers seriously ...
详细信息
Recent works have shown the remarkable superiority of transformer models in reinforcement learning (RL), where the decision-making problem is formulated as sequential generation. Transformer-based agents could emerge ...
ISBN:
(纸本)9798331314385
Recent works have shown the remarkable superiority of transformer models in reinforcement learning (RL), where the decision-making problem is formulated as sequential generation. Transformer-based agents could emerge with self-improvement in online environments by providing task contexts, such as multiple trajectories, called in-context RL. However, due to the quadratic computation complexity of attention in transformers, current in-context RL methods suffer from huge computational costs as the task horizon increases. In contrast, the Mamba model is renowned for its efficient ability to process long-term dependencies, which provides an opportunity for in-context RL to solve tasks that require long-term memory. To this end, we first implement Decision Mamba (DM) by replacing the backbone of Decision Transformer (DT). Then, we propose a Decision Mamba-Hybrid (DM-H) with the merits of transformers and Mamba in high-quality prediction and long-term memory. Specifically, DM-H first generates high-value sub-goals from long-term memory through the Mamba model. Then, we use sub-goals to prompt the transformer, establishing high-quality predictions. Experimental results demonstrate that DM-H achieves state-of-the-art in long and short-term tasks, such as D4RL, Grid World, and Tmaze benchmarks. Regarding efficiency, the online testing of DM-H in the long-term task is 28× times faster than the transformer-based baselines.
Combination therapy is a promising approach to address the challenge of antimicrobial resistance,and computational models have been proposed for predicting drug–drug *** existing models rely on drug similarity measur...
详细信息
Combination therapy is a promising approach to address the challenge of antimicrobial resistance,and computational models have been proposed for predicting drug–drug *** existing models rely on drug similarity measures based on characteristics such as chemical structure and the mechanism of *** this study,we focus on the network structure itself and propose a drug similarity measure based on drug–drug interaction *** explore the potential applications of this measure by combining it with unsupervised learning and semi-supervised learning *** unsupervised learning,drugs can be grouped based on their interactions,leading to almost monochromatic group–group *** addition,drugs within the same group tend to have similar mechanisms of action(MoA).In semi-supervised learning,the similarity measure can be utilized to construct affinity matrices,enabling the prediction of unknown drug–drug *** method exceeds existing approaches in terms of ***,our experiments demonstrate the effectiveness and practicability of the proposed similarity *** the one hand,when combined with clustering algorithms,it can be used for functional annotation of compounds with unknown *** the other hand,when combined with semi-supervised graph learning,it enables the prediction of unknown drug–drug interactions.
In online insurance, one of the central challenges is the cold-starting of new insurance products, which means there are no previous samples to refer to. Previous studies have mainly focused on improving the predictio...
暂无评论