With the proliferation of mobile sensing techniques, huge amounts of time series data are generated and accumulated in various domains, fueling plenty of real-world applications. In this setting, time series anomaly detection is practically important. It endeavors to identify deviant samples from the normal sample distribution in time series. Existing approaches generally assume that all the time series is available at a central location. However, we are witnessing the decentralized collection of time series due to the deployment of various edge devices. To bridge the gap between the decentralized time series data and the centralized anomaly detection algorithms, we propose a Parameter-efficient Federated Anomaly Detection framework named PeFAD with the increasing privacy concerns. PeFAD for the first time employs the pre-trained language model (PLM) as the body of the client's local model, which can benefit from its cross-modality knowledge transfer capability. To reduce the communication overhead and local model adaptation cost, we propose a parameter-efficient federated training module such that clients only need to fine-tune small-scale parameters and transmit them to the server for update. PeFAD utilizes a novel anomaly-driven mask selection strategy to mitigate the impact of neglected anomalies during training. A knowledge distillation operation on a synthetic privacy-preserving dataset that is shared by all the clients is also proposed to address the data heterogeneity issue across clients. We conduct extensive evaluations on four real datasets, where PeFAD outperforms existing state-of-the-art baselines by up to 28.74%.

Consider a reinforcement learning problem where an agent has access to a very large amount of information about the environment, but it can only take very few actions to accomplish its task and to maximize its reward. Evidently, the main problem for the agent is to learn a map from a very high-dimensional space (which represents its environment) to a very low-dimensional space (which represents its actions). The high-to-low dimensional map implies that most of the information about the environment is irrelevant for the actions to be taken, and only a small fraction of information is relevant. In this paper we argue that the relevant information need not be learned by brute force (which is the standard approach), but can be identified from the intrinsic symmetries of the system. We analyze in details a reinforcement learning problem of autonomous driving, where the corresponding symmetry is the Galilean symmetry, and argue that the learning task can be accomplished with very few relevant parameters, or, more precisely, invariants. For a numerical demonstration, we show that the autonomous vehicles (which we call autonomous particles since they describe very primitive vehicles) need only four relevant invariants to learn how to drive very well without colliding with other particles. The simple model can be easily generalized to include different types of particles (e.g. for cars, for pedestrians, for buildings, for road signs, etc.) with different types of relevant invariants describing interactions between them. We also argue that there must exist a field theory description of the learning system where autonomous particles would be described by fermionic degrees of freedom and interactions mediated by the relevant invariants would be described by bosonic degrees of freedom. This suggests that the effectiveness of field theory descriptions of physical systems might be connected to the learning dynamics of some kinds of autonomous particles, supporting the claim that the entire universe is a neural network.

SHANGHAI, Sept 1 (Reuters) - A U.S. order to ban exports of some advanced chips to China is likely to hit almost any major tech company running public clouds or advanced artificial intelligence training modules in the country, experts said. Chip designer Nvidia Corp (NVDA.O) said on Wednesday that U.S. officials told it to stop exporting two top computing chips for AI work to China. Advanced Micro Devices (AMD.O) also said it had received new license requirements that will stop its advanced AI chip called MI250 from being exported to China. Shu Jueting, a Chinese Commerce Ministry spokesperson, said on Thursday that Beijing opposes the measures, saying they undermine the rights of Chinese companies and threaten to disrupt global supply chains. The orders underscore deepening U.S.-China tensions over access to advanced chip technology.

Artificial intelligence is evolving the landscape of every industry, and the gaming industry is no exception. Through innovation and growth, technology is exceeding our expectations every day. In gaming, artificial intelligence (AI) refers to responsive and flexible video game experiences. While artificial intelligence has long been present in video games, it is today seen as a burgeoning new frontier in how games are both created and played. AI games are progressively handing over control of the game experience to the player, whose actions influence the game experience.

Artificial intelligence has been a great influencer in almost every industry these days. There are a number of advantages in terms of resource efficiency, resource optimisation, availability and high accuracy to name a few. We all have benefited from AI in some way, shape or form and AI will keep impacting our lives in a positive manner for the rest of our lives. While AI is here to stay with its advantages, I have been speculating use of Artificial Intelligence across domains and have been curious about the various applications. It all started with Cambridge Analytica documentary (CA) where CA team allegedly used AI to target voters on the edge to shift the US presidential election dynamics in 2016.

In September 2016, Salesforce founder and CEO Marc Benioff informed employees, customers, and investors that Salesforce would be an AI-driven company. Earlier that year, Microsoft released its Tay research chatbot project through a Twitter Account. Microsoft shut down Tay after only 16 hours because it started to mimic the deliberately offensive behavior of other Twitter users, and Microsoft had not given the bot an understanding of such inappropriate behavior. With chatbots as one of Salesforce's most promising customer service-related technologies, Kathy Baxter, in her role at that time as Principal User Researcher, was curious about understanding what went wrong with Tay. She also wanted to know how that type of AI-enabled system behavior could be avoided at Salesforce.

Despite the improvement of translation quality, neural machine translation (NMT) often suffers from the lack of diversity in its generation. In this paper, we propose to generate diverse translations by deriving a large number of possible models with Bayesian modelling and sampling models from them for inference. The possible models are obtained by applying concrete dropout to the NMT model and each of them has specific confidence for its prediction, which corresponds to a posterior model distribution under specific training data in the principle of Bayesian modeling. With variational inference, the posterior model distribution can be approximated with a variational distribution, from which the final models for inference are sampled. We conducted experiments on Chinese-English and English-German translation tasks and the results shows that our method makes a better trade-off between diversity and accuracy.

Artificial Intelligence (AI) could increase global GDP by 14 percent, or an astounding $15.7 trillion by 2030. This is due, in large part, to productivity gains from AI automation and workforce augmentation. AI will change the world, but it takes time to implement and train it. It's important for your business to understand how, when, and where to implement Artificial Intelligence, and it's often best to start small. The world at large is still learning how best AI can be used to benefit society.

Earlier this week, Google announced that it was piloting a machine learning intensive for college students. Today, its broader Machine Learning Crash Course is adding a new training module on fairness when building AI. As adoption of machine learning continues, ethics and fairness are very important considerations. While AI can have the "potential to be fairer and more inclusive at a broader scale than decision-making processes based on ad hoc rules or human judgments," there might be underlying biases present in the data used to train these models. Other issues involve insuring that AI is fair in all situations, while more broadly there is "no standard definition of fairness."

I am from Kashmir, a region prone to earthquakes and floods. When I was 17 years old, in 2005, 70,000 people lost their lives in an earthquake in my hometown. This event compelled me to study engineering and specifically in 2005, start performing earthquake engineering research. Then, in 2014, a combination of two events on different sides of the world inspired the creation of One Concern. In 2014, during a break from graduate school at Stanford, I was visiting my parents in Kashmir when a large flood engulfed the state.