Graph Neural Networks (GNNs) are non-Euclidean deep learning models for graph-structured data.

Graph Neural Networks (GNNs) are non-Euclidean deep learning models for graph-structured data.

Graph Neural Networks (GNNs) are non-Euclidean deep learning models for graph-structured data. Despite their successful and diverse applications, oversmoothing prohibits deep architectures due to node features converging to a single fixed point. This severely limits their potential to solve complex tasks. To counteract this tendency, we propose a plug-and-play module consisting of three steps: Cluster-Normalize-Activate (CNA). By applying CNA modules, GNNs search and form super nodes in each layer, which are normalized and activated individually. We demonstrate in node classification and property prediction tasks that CNA significantly improves the accuracy over the state-of-the-art. Particularly, CNA reaches 94.18% and 95.75% accuracy on Cora and CiteSeer, respectively. It further benefits GNNs in regression tasks as well, reducing the mean squared error compared to all baselines. At the same time, GNNs with CNA require substantially fewer learnable parameters than competing architectures.

This paper considers planning a path for a single underwater cooperative navigation aid (CNA) vehicle to sequentially aid a set of N agents to minimize average navigation uncertainty. Both the CNA and agents are modeled as constant-velocity vehicles. The agents travel along known nominal trajectories and the CNA plans a path to sequentially intercept them. Navigation aiding is modeled by a scalar discrete time Kalman filter. During path planning, the CNA considers surfacing to reduce its own navigation uncertainty. A greedy planning algorithm is proposed that uses a heuristic to schedule agents to the CNA that is based on the optimal time-to-aid, the overall navigation uncertainty reduction, and the transit time. The approach is compared to an optimal (exhaustive enumeration) algorithm through a Monte Carlo experiment with randomized agent trajectories and initial navigation uncertainty.

Neural Architecture Search (NAS) paves the way for the automatic definition of Neural Network (NN) architectures, attracting increasing research attention and offering solutions in various scenarios. This study introduces a novel NAS solution, called Flat Neural Architecture Search (FlatNAS), which explores the interplay between a novel figure of merit based on robustness to weight perturbations and single NN optimization with Sharpness-Aware Minimization (SAM). FlatNAS is the first work in the literature to systematically explore flat regions in the loss landscape of NNs in a NAS procedure, while jointly optimizing their performance on in-distribution data, their out-of-distribution (OOD) robustness, and constraining the number of parameters in their architecture. Differently from current studies primarily concentrating on OOD algorithms, FlatNAS successfully evaluates the impact of NN architectures on OOD robustness, a crucial aspect in real-world applications of machine and deep learning. FlatNAS achieves a good trade-off between performance, OOD generalization, and the number of parameters, by using only in-distribution data in the NAS exploration. The OOD robustness of the NAS-designed models is evaluated by focusing on robustness to input data corruptions, using popular benchmark datasets in the literature.

Even the cleverest, most cunning artificial intelligence algorithm will presumably have to obey the laws of silicon. Its capabilities will be constrained by the hardware that it's running on. Some researchers are exploring ways to exploit that connection to limit the potential of AI systems to cause harm. The idea is to encode rules governing the training and deployment of advanced algorithms directly into the computer chips needed to run them. In theory--the sphere where much debate about dangerously powerful AI currently resides--this might provide a powerful new way to prevent rogue nations or irresponsible companies from secretly developing dangerous AI.

Perhaps without realising it, recruiters are already using AI and enjoying the benefits of it. Take LinkedIn for example, where its recruiter-facing product will automatically suggest top candidates based on keywords or other parameters we feed it. There are countless examples of AI doing background work for busy talent acquisition professionals, letting them focus on what is important. The reason that we are seeing more conversations around ChatGPT and generative AI specifically is because of how in-your-face it is. Perhaps ignorance has been bliss in the past?

SINGAPORE: Who do you think should be responsible when artificial intelligence or algorithms malfunction: The programmer, manufacturer or user? Singapore plans to be a global leader in artificial intelligence (AI) by 2030. This involves, on the one hand, widespread deployment of AI in a variety of settings, and on the other, widespread trust in these AI solutions. Clearly that trust needs to be well-placed, but what does it mean for trust to be well-placed? Certainly, one part of this is AI getting things right reliably often. But that alone is not enough.

Amid swirling predictions of how an artificial intelligence chatbot could … ChatGPT is nothing more than a sophisticated machine-learning system …

This study attempts to investigate cross-strait variations on two typical synonymous loanwords in Chinese, i.e. xie2shang1 and tan2pan4, drawn on MARVS theory. Through a comparative analysis, the study found some distributional, eventual, and contextual similarities and differences across Taiwan and Mainland Mandarin. Compared with the underused tan2pan4, xie2shang1 is significantly overused in Taiwan Mandarin and vice versa in Mainland Mandarin. Additionally, though both words can refer to an inchoative process in Mainland and Taiwan Mandarin, the starting point for xie2shang1 in Mainland Mandarin is somewhat blurring compared with the usage in Taiwan Mandarin. Further on, in Taiwan Mandarin, tan2pan4 can be used in economic and diplomatic contexts, while xie2shang1 is used almost exclusively in political contexts. In Mainland Mandarin, however, the two words can be used in a hybrid manner within political contexts; moreover, tan2pan4 is prominently used in diplomatic contexts with less reference to economic activities, while xie2sahng1 can be found in both political and legal contexts, emphasizing a role of mediation.