As deep learning models become increasingly large, they pose significant challenges in heterogeneous devices environments. The size of deep learning models makes it difficult to deploy them on low-power or resource-constrained devices, leading to long inference times and high energy consumption. To address these challenges, we propose FlexTrain, a framework that accommodates the diverse storage and computational resources available on different devices during the training phase. FlexTrain enables efficient deployment of deep learning models, while respecting device constraints, minimizing communication costs, and ensuring seamless integration with diverse devices. We demonstrate the effectiveness of FlexTrain on the CIFAR-100 dataset, where a single global model trained with FlexTrain can be easily deployed on heterogeneous devices, saving training time and energy consumption. We also extend FlexTrain to the federated learning setting, showing that our approach outperforms standard federated learning benchmarks on both CIFAR-10 and CIFAR-100 datasets.

With recent advances in sensing technologies, a myriad of spatio-temporal data has been generated and recorded in smart cities. Forecasting the evolution patterns of spatio-temporal data is an important yet demanding aspect of urban computing, which can enhance intelligent management decisions in various fields, including transportation, environment, climate, public safety, healthcare, and others. Traditional statistical and deep learning methods struggle to capture complex correlations in urban spatio-temporal data. To this end, Spatio-Temporal Graph Neural Networks (STGNN) have been proposed, achieving great promise in recent years. STGNNs enable the extraction of complex spatio-temporal dependencies by integrating graph neural networks (GNNs) and various temporal learning methods. In this manuscript, we provide a comprehensive survey on recent progress on STGNN technologies for predictive learning in urban computing. Firstly, we provide a brief introduction to the construction methods of spatio-temporal graph data and the prevalent deep-learning architectures used in STGNNs. We then sort out the primary application domains and specific predictive learning tasks based on existing literature. Afterward, we scrutinize the design of STGNNs and their combination with some advanced technologies in recent years. Finally, we conclude the limitations of existing research and suggest potential directions for future work.

In response to anthropogenic climate change, many countries and international organisations have committed to legally binding greenhouse gas emissions targets. The UK and the EU have both recently updated their legislation to include net zero emissions targets in place for 2050 (Skidmore, 2019; Sassoli and Matos Fernandes, 2021). This requires moving away from using fossil fuels for energy generation and moving towards renewable sources such as photovoltaic cells and wind turbines. Centralised'national grids' are able to'switch on and off' traditional fossil fuel power plants in order to increase or decrease the energy supply to meet the demand of the users. As the proportion of energy being generated from renewable sources increases this raises a problem - how can load-balancing (the matching of supply and demand) be managed when the output is inherently dependent on weather conditions. This load-balancing problem is easier to address on a small scale, and as such governments and energy providers are supporting the development of'Community energy systems', where local communities such as a small town own and manage their own renewable energy resources (Walker and Devine-Wright, 2008; Gruber et al., 2021). Decentralised community energy systems allow for a higher share of renewable technologies to be integrated into energy generation (Chiradeja and Ramakumar, 2004); minimise transmission losses between the source of energy generation and the end users (Pepermans et al., 2005); and improve energy security as the energy supply is less impacted by geopolitical factors (Alanne and Saari, 2006). As social awareness of environmental issues increases, the willingness of communities to invest in community energy systems is also expected to increase (Pasimeni, 2019). While there are clear benefits to widespread adoption, the shift towards community energy systems means that comarXiv:2211.10198v2

A networked time series (NETS) is a family of time series on a given graph, one for each node. It has a wide range of applications from intelligent transportation, environment monitoring to smart grid management. An important task in such applications is to predict the future values of a NETS based on its historical values and the underlying graph. Most existing methods require complete data for training. However, in real-world scenarios, it is not uncommon to have missing data due to sensor malfunction, incomplete sensing coverage, etc. In this paper, we study the problem of NETS prediction with incomplete data. We propose NETS-ImpGAN, a novel deep learning framework that can be trained on incomplete data with missing values in both history and future. Furthermore, we propose Graph Temporal Attention Networks, which incorporate the attention mechanism to capture both inter-time series and temporal correlations. We conduct extensive experiments on four real-world datasets under different missing patterns and missing rates. The experimental results show that NETS-ImpGAN outperforms existing methods, reducing the MAE by up to 25%.

The number of sampling methods could be daunting for a practitioner looking to cast powerful machine learning methods to their specific problem. This paper takes a theoretical stance to review and organize many sampling approaches in the ``generative modeling'' setting, where one wants to generate new data that are similar to some training examples. By revealing links between existing methods, it might prove useful to overcome some of the current challenges in sampling with diffusion models, such as long inference time due to diffusion simulation, or the lack of diversity in generated samples.

Back in the 1990s, anyone suggesting that we'd need to adapt to climate change while also cutting emissions was met with suspicion. Most climate change researchers felt adaptation studies would distract from the vital work of keeping pollution out of the atmosphere to begin with. Despite this hostile environment, a handful of experts were already sowing the seeds for a new field of research called "climate change adaptation": study and policy on how the world could prepare for and adapt to the new disasters and dangers brought forth on a warming planet. Today, their research is more important than ever. Take a moment out of your day to appreciate these truly remarkable rivers.

The US Department of Energy has recognized this trend, recently awarding $3 billion in grants to various "smart grid" projects that include AI-related initiatives. The excitement about AI in the energy sector is palpable. Some are already speculating about the possibility of a fully automated grid where, in theory, no humans would be needed to make everyday decisions. But that prospect remains far off; for now, the promise lies in the potential for AI to help humans, providing real-time insights for better grid management. Here are four of the ways that AI is already changing how grid operators do their work.

The simulation of geological facies in an unobservable volume is essential in various geoscience applications. Given the complexity of the problem, deep generative learning is a promising approach to overcome the limitations of traditional geostatistical simulation models, in particular their lack of physical realism. This research aims to investigate the application of generative adversarial networks and deep variational inference for conditionally simulating meandering channels in underground volumes. In this paper, we review the generative deep learning approaches, in particular the adversarial ones and the stabilization techniques that aim to facilitate their training. The proposed approach is tested on 2D and 3D simulations generated by the stochastic process-based model Flumy. Morphological metrics are utilized to compare our proposed method with earlier iterations of generative adversarial networks. The results indicate that by utilizing recent stabilization techniques, generative adversarial networks can efficiently sample from target data distributions. Moreover, we demonstrate the ability to simulate conditioned simulations through the latent variable model property of the proposed approach.

In highly interactive driving scenarios, the actions of one agent greatly influences those of its neighbors. Planning safe motions for autonomous vehicles in such interactive environments, therefore, requires reasoning about the impact of the ego's intended motion plan on nearby agents' behavior. Deep-learning-based models have recently achieved great success in trajectory prediction and many models in the literature allow for ego-conditioned prediction. However, leveraging ego-conditioned prediction remains challenging in downstream planning due to the complex nature of neural networks, limiting the planner structure to simple ones, e.g., sampling-based planner. Despite their ability to generate fine-grained high-quality motion plans, it is difficult for gradient-based planning algorithms, such as model predictive control (MPC), to leverage ego-conditioned prediction due to their iterative nature and need for gradient. We present Interactive Joint Planning (IJP) that bridges MPC with learned prediction models in a computationally scalable manner to provide us the best of both the worlds. In particular, IJP jointly optimizes over the behavior of the ego and the surrounding agents and leverages deep-learned prediction models as prediction priors that the join trajectory optimization tries to stay close to. Furthermore, by leveraging homotopy classes, our joint optimizer searches over diverse motion plans to avoid getting stuck at local minima. Closed-loop simulation result shows that IJP significantly outperforms the baselines that are either without joint optimization or running sampling-based planning.

We introduce PhysGaussian, a new method that seamlessly integrates physically grounded Newtonian dynamics within 3D Gaussians to achieve high-quality novel motion synthesis. Employing a custom Material Point Method (MPM), our approach enriches 3D Gaussian kernels with physically meaningful kinematic deformation and mechanical stress attributes, all evolved in line with continuum mechanics principles. A defining characteristic of our method is the seamless integration between physical simulation and visual rendering: both components utilize the same 3D Gaussian kernels as their discrete representations. This negates the necessity for triangle/tetrahedron meshing, marching cubes, "cage meshes," or any other geometry embedding, highlighting the principle of "what you see is what you simulate (WS$^2$)." Our method demonstrates exceptional versatility across a wide variety of materials--including elastic entities, metals, non-Newtonian fluids, and granular materials--showcasing its strong capabilities in creating diverse visual content with novel viewpoints and movements. Our project page is at: https://xpandora.github.io/PhysGaussian/