Generative Adversarial Networks (GANs) have promoted a variety of applications in computer vision and natural language processing, among others, due to its generative model’s compelling ability to generate realistic examples plausibly drawn from an existing distribution of samples. GAN not only provides impressive performance on data generation-based tasks but also stimulates fertilization for privacy and security oriented research because of its game theoretic optimization strategy. Unfortunately, there are no comprehensive surveys on GAN in privacy and security, which motivates this survey to summarize systematically. The existing works are classified into proper categories based on privacy and security functions, and this survey conducts a comprehensive analysis of their advantages and drawbacks. Considering that GAN in privacy and security is still at a very initial stage and has imposed unique challenges that are yet to be well addressed, this article also sheds light on some potential privacy and security applications with GAN and elaborates on some future research directions.

https://dl.acm.org/doi/pdf/10.1145/3459992

Generative Adversarial Networks: A Survey Toward Private and Secure Applications

Quantitative assessment of U.S. bulk power systems and market operations during the COVID-19 pandemic

According to the World Health Organization (WHO), the COVID-19 pandemic has been the greatest challenge to humanity since World War II Spain is one of the countries experiencing severe consequences from the COVID-19 pandemic outbreak In such a situation, when the population is quarantined, hospitals are crowded, and industries are shutting down;energy supply becomes a critical and “hard to predict” service In this way, the study of the impact of COVID-19 on the selling price of electricity in Spain is highlighted more comprehensively Given the high level of uncertainty in global energy markets, there is a research gap that needs to be filled, that is, the creation of new decision support mechanisms that can reduce risks and ensure more significant forecasting for investments and energy management to deal with seemingly unprecedented effects experienced through the current pandemic This study examines and develops econometric models to determine the impact of COVID-19 on the Spanish electricity market This article's main result is a framework that predicts the dynamic electricity market's behavior during shifting periods, that is, the COVID-19 pandemic period Noteworthy that the result of this article is useful in electricity market management and strategic developments © 2020 John Wiley & Sons Ltd

The impact of COVID-19 on the electricity sector in Spain: An econometric approach based on prices

Standard federated optimization methods successfully apply to stochastic problems with single-level structure. However, many contemporary ML problems -- including adversarial robustness, hyperparameter tuning, and actor-critic -- fall under nested bilevel programming that subsumes minimax and compositional optimization. In this work, we propose \fedblo: A federated alternating stochastic gradient method to address general nested problems. We establish provable convergence rates for \fedblo in the presence of heterogeneous data and introduce variations for bilevel, minimax, and compositional optimization. \fedblo introduces multiple innovations including federated hypergradient computation and variance reduction to address inner-level heterogeneity. We complement our theory with experiments on hyperparameter \&hyper-representation learning and minimax optimization that demonstrate the benefits of our method in practice. Code is available at https://github.com/ucr-optml/FedNest.

FEDNEST: Federated Bilevel, Minimax, and Compositional Optimization

To contain the spread of the novel coronavirus (COVID-19), local and state governments in the U.S. have imposed restrictions on daily life, resulting in dramatic changes to how and where people interact, travel, socialize, and work. Using a social practice perspective, we explore how California's Shelter-in-Place (SIP) order impacted household energy activities. To do so, we conducted an online survey of California residents (n = 804) during active SIP restrictions (May 5–18, 2020). We asked respondents about changes to home occupancy patterns and household energy activities (e.g., cooking, electronics usage) due to SIP restrictions, as well as perspectives toward smart energy technologies. Households reported increased midday (10am–3pm) occupancy during SIP, and this increase is related to respondent and household characteristics, such as education and the presence of minors in the home. Examining change in the frequency of household activities during SIP, presence of minors and increased midday occupancy proved important. Finally, we considered relationships to intention to purchase smart home technologies, with the presence of minors and increased activity frequency relating to greater intention to purchase. These findings demonstrate how household activities and occupancy changed under COVID restrictions, how these changes may be related to energy use in the home, and how such COVID-related changes could be shaping perspectives toward smart home technology, potentially providing insight into future impacts on household practices and electricity demand.

Exploring the effects of California's COVID-19 shelter-in-place order on household energy practices and intention to adopt smart home technologies

We propose Federated Generative Adversarial Network (FedGAN) for training a GAN across distributed sources of non-independent-and-identically-distributed data sources subject to communication and privacy constraints. Our algorithm uses local generators and discriminators which are periodically synced via an intermediary that averages and broadcasts the generator and discriminator parameters. We theoretically prove the convergence of FedGAN with both equal and two time-scale updates of generator and discriminator, under standard assumptions, using stochastic approximations and communication efficient stochastic gradient descents. We experiment FedGAN on toy examples (2D system, mixed Gaussian, and Swiss role), image datasets (MNIST, CIFAR-10, and CelebA), and time series datasets (household electricity consumption and electric vehicle charging sessions). We show FedGAN converges and has similar performance to general distributed GAN, while reduces communication complexity. We also show its robustness to reduced communications.

FedGAN: Federated Generative Adversarial Networks for Distributed Data.

Understanding how efforts to limit exposure to COVID-19 have altered electricity demand provides insights not only into how dramatic restrictions shape electricity demand but also about future electricity use in a post-COVID-19 world. We develop a unified modeling framework to quantify and compare electricity usage changes in 58 countries and regions around the world from January-May 2020. We find that daily electricity demand declined as much as 10% in April 2020 compared to modelled demand, controlling for weather, seasonal and temporal effects, but with significant variation. Clustering techniques show that four impact groups capture systematic differences in timing and depth of electricity usage changes, ranging from a mild decline of 2% to an extreme decline of 26%. These groupings do not align with geography, with almost every continent having at least one country or region that experienced a dramatic reduction in demand and one that did not. Instead, we find that such changes relate to government restrictions and mobility. Government restrictions have a non-linear effect on demand that generally saturates at its most restrictive levels and sustains even as restrictions ease. Mobility offers a sharper focus on electricity demand change with workplace and residential mobility strongly linked to demand changes at the daily level. Steep declines in electricity usage are associated with workday hourly load patterns that resemble pre-COVID weekend usage. Quantifying these impacts is a crucial first step in understanding the impacts of crises like the pandemic and the associated societal response on electricity demand.

Power and the Pandemic: Exploring Global Changes in Electricity Demand During COVID-19.

ABSTRACT The COVID-19 outbreak has the world gripped by fear and panic with its high velocity of infection. Based on the first three months of the global pandemic centred in Asia, this study investigates the relationship between the perception of COVID-19 on consumers’ destination image towards Wuhan and China, and how risk perceptions and changes in destination image affect travel intention to the destinations. Findings illuminated that perceptions of COVID-19 do not have a direct effect on destination image. Instead, risk perception of travelling during the pandemic mediated its effects on the destination image of Wuhan and China. Theoretical and practical implications concerning the management of destination image for the recovery of tourism are discussed with future possibilities of this research.

Risk and destination perceptions of Wuhan, China since the COVID-19 pandemic

The timing of electricity consumption is increasingly important for grid operations. In response, households are being encouraged to alter their daily usage patterns through demand response and time-varying pricing, although it is unknown if they are aware of these patterns. Here we introduce an energy literacy concept, 'load shape awareness', and apply it to a sample of California residents (n = 186) who provided their household’s hourly electricity data and completed an energy use questionnaire. Choosing from four prominent load shape designations, half of respondents (51%) correctly identified their dominant load shape before COVID-19 shelter-in-place (SIP) orders while only one-third (31%) did so during SIP orders. Those aware of their load shape were more likely to have chosen evening peak, the most frequent dominant shape in the electricity data. Our work provides proof of principle for the load shape awareness concept, which could prove useful in designing energy conservation interventions and helping consumers adapt to an evolving energy system. How well households understand their own energy usage patterns may impact the effectiveness of demand response initiatives. Zanocco et al. find that only half of a sample of California households were able to identify their own usage pattern from among four test patterns before COVID-19 restrictions. 

/pdf/assessing-californians-awareness-of-their-daily-electricity-vxlddezw.pdf

Assessing Californians’ awareness of their daily electricity use patterns

The electric grid is uniquely susceptible to extreme events through both power supply and consumption pathways. Extreme events - like heatwaves and droughts - are expected to increase in frequency and severity due to climate change and are already causing consequences on power system operations and stability. Additionally, non-climate related events like the COVID-19 pandemic have had dramatic impacts on energy consumption patterns globally. We apply modern machine learning methods to model electricity consumption in Brazil, one of the largest generators of hydropower, to better understand the consumption-side effects of extreme national and regional events. After training on 20 years of historical data, we verify an R2of 0.848 and a MAPE of 2.6% for our counterfactual model and use it to assess impacts of historical events on electricity consumption. We then discuss how this approach can be applied toward measuring energy system responsiveness and resiliency on present and future scenarios.

Tao Sun

Papers

FedGAN: Federated Generative Adversarial Networks for Distributed Data.

Power and the Pandemic: Exploring Global Changes in Electricity Demand During COVID-19.

Risk and destination perceptions of Wuhan, China since the COVID-19 pandemic

Assessing Californians’ awareness of their daily electricity use patterns

Revealing the Impact of Extreme Events on Electricity Consumption in Brazil: A Data-Driven Counterfactual Approach