Increasing energy consumption and diversified energy generation have led to the proliferation of energy management (EM) systems for optimized grid operations and net zero carbon emissions. Artificial Intelligence (AI)...
Increasing energy consumption and diversified energy generation have led to the proliferation of energy management (EM) systems for optimized grid operations and net zero carbon emissions. Artificial Intelligence (AI) is centric to EM systems to process large volumes of high velocity data for anomalies, predictions, optimization and other actionable insights that unravel the complexities of grid operations. Managing Energy AI capabilities itself is becoming an increasingly complex task that requires extensive resourcing and expertise. In this paper, we aim to address this gap by formalizing the role of Automated Machine Learning (AutoML) by proposing a novel framework for its key functionalities in critical energy infrastructure. This framework provides a generic and cohesive abstraction to assist with integrating and managing the complexities of AI capabilities. The framework is empirically evaluated in the microgrid setting of a multi-campus, multi-functional tertiary education institution. Results from these experiments confirm the performance contributions of the proposed framework in addressing the complexities of AI capabilities of EM systems as they transition towards microgrid energy optimisation and net zero carbon emissions.
With the development of society and economy, the number of cars on the road has exploded and caused a series of traffic problems. Only relying on the rear-view mirror has been unable to meet the safety requirements of...
详细信息
Affective interaction with virtual humans can enhance the quality of user experience in virtual reality. It takes place through various considerations such as emotional representation in their behavioral patterns, fac...
详细信息
ISBN:
(纸本)9781665457262
Affective interaction with virtual humans can enhance the quality of user experience in virtual reality. It takes place through various considerations such as emotional representation in their behavioral patterns, facial expressions, head pose, body stance, and so on. Deciding on the emotional state of a virtual human at a moment, however, is still a challenge. Computational models of emotion, stemming from appraisal theories, are suggested for modeling emotion in virtual agents. Despite their competence in extracting emotion from appraisal values, they are poorly defined in describing how to assess appraisal values within a sense-think-act behavior model. Motivated by this lack of empirical knowledge on the appraisal stage, in this preliminary work, we propose a framework to bridge the gap between a computational model of emotion and a behavior model of virtual humans. To this end, we use a need-based, goal- oriented, autonomous behavior model to generate salient stimuli for eliciting emotions. Our simulation of a case study suggests that the proposed framework can produce sensible emotional states that conform to the essential principles of appraisal-based emotion theories.
Two novel silica-based hybrid materials, M1 and M2, based on silica gel and MCM-41 with a ditopic triazole-pyrazole ligand grafted onto their surfaces, respectively, were successfully synthesized and fully characteriz...
详细信息
The detection of anomalies in multivariate time series data is crucial for various practical applications, including smart power grids, traffic flow forecasting, and industrial process control. However, real-world tim...
详细信息
Smart contracts and the blockchain have recently been widely used in many application fields. Current smart contracts are executed on general-purpose CPUs and still have a large room to improve performance. In this pa...
Smart contracts and the blockchain have recently been widely used in many application fields. Current smart contracts are executed on general-purpose CPUs and still have a large room to improve performance. In this paper, we first analyze the most popular public blockchain platform Ethereum and characterize smart contracts running on its ecosystem. After identifying its performance limitations, we propose a heterogeneous processor Smart Contract Unit (SCU), which is a hardware-based accelerator in place of the current EVM design. With our proposed novel RISC-style SCU ISA and heterogeneous architecture, SCU can leverage instruction-level parallelism and transaction-level parallelism during smart contract processing and boost its execution performance. Furthermore, SCU can be configured and adapted to different workloads in order to remove bottlenecks. We implement and evaluate the proposed SCU design on a Xilinx FPGA platform. Our design achieves a significant speedup compared to the software implementation on an Intel CPU and runs a few times faster than state-of-the-art design.
With ever increasing amounts of travel, it is essential to have access to a patient’s medical data from different sources including many jurisdictions. The Serums project addresses this goal by creating a healthcare ...
详细信息
Depth-3 circuit lower bounds and k-SAT algorithms are intimately related;the state-of-the-art Σk3 -circuit lower bound (Or-And-Or circuits with bottom fan-in at most k) and the k-SAT algorithm of Paturi, Pudlák,...
详细信息
ISBN:
(纸本)9783959773317
Depth-3 circuit lower bounds and k-SAT algorithms are intimately related;the state-of-the-art Σk3 -circuit lower bound (Or-And-Or circuits with bottom fan-in at most k) and the k-SAT algorithm of Paturi, Pudlák, Saks, and Zane (J. ACM’05) are based on the same combinatorial theorem regarding k-CNFs. In this paper we define a problem which reveals new interactions between the two, and suggests a concrete approach to significantly stronger circuit lower bounds and improved k-SAT algorithms. For a natural number k and a parameter t, we consider the Enum(k, t) problem defined as follows: given an n-variable k-CNF and an initial assignment α, output all satisfying assignments at Hamming distance t(n) of α, assuming that there are no satisfying assignments of Hamming distance less than t(n) of α. We observe that an upper bound b(n, k, t) on the complexity of Enum(k, t) simultaneously implies depth-3 circuit lower bounds and k-SAT algorithms: Depth-3 circuits: Any Σk3 circuit computing the Majority function has size at least (nn2 )/b(n, k, n2 ). k-SAT: There exists an algorithm solving k-SAT in time O (Pn/t=12 b(n, k, t) ) . A simple construction shows that b(n, k, n2 ) ≥ 2(1−O(log(k)/k))n. Thus, matching upper bounds for b(n, k, n2 ) would imply a Σk3 -circuit lower bound of 2Ω(log(k)n/k) and a k-SAT _upper bound of 2(1−Ω(log(k)/k))n. The former yields an unrestricted depth-3 lower bound of 2ω(√n) solving a long standing open problem, and the latter breaks the Super Strong Exponential Time Hypothesis. In this paper, we propose a randomized algorithm for Enum(k, t) and introduce new ideas to analyze it. We demonstrate the power of our ideas by considering the first non-trivial instance of the problem, i.e., Enum(3, n2 ). We show that the expected running time of our algorithm is 1.598n, substantially improving on the trivial bound of 3n/2 1.732n. This already improves Σ33 lower bounds for Majority function to 1.251n. The previous bound was 1.154n which follows from the wor
Machine perception of emotions is integral to the development of human-centric Artificial Intelligence (AI) in sustainable industrial applications. Human expressions of emotions are not always direct. Word embeddings ...
Machine perception of emotions is integral to the development of human-centric Artificial Intelligence (AI) in sustainable industrial applications. Human expressions of emotions are not always direct. Word embeddings are mature techniques that can extract the semantics of such indirect expressions from text data. However, they are not primed to extract emotions. In this paper, we propose a novel approach that generates robust word embeddings for implicit and explicit expressions of emotion. This approach consists of two techniques, mask and rogue, we evaluate both techniques on two benchmark datasets for emotion classification. Our results confirm the effectiveness of the proposed approach in extracting emotions from diverse contexts. We have shared the emotion word embedding for public use.
The Internet of Things (IoT) is becoming increasingly ubiquitous, typified in software by large-scale multi-agent systems of heterogeneous agents. IoT devices are constrained in terms of memory and processing power, l...
The Internet of Things (IoT) is becoming increasingly ubiquitous, typified in software by large-scale multi-agent systems of heterogeneous agents. IoT devices are constrained in terms of memory and processing power, limiting their capacity to hold large sets of information upon which decision-making logic must execute. IoT devices are also frequently deployed as distributed sensors constrained in terms of time, location and communications bandwidth. These constraints demand a level of multi-agent communication and co-ordination to provide accurate, up-to-date information on-demand to different intelligent agents in the system. In this paper, we propose a novel method by which Aspect-Oriented Software Design can be applied to managing the validity of time-constrained data in IoT systems while decoupling the application code from this concern.
暂无评论