Deep reinforcement learning (RL) agents may successfully generalize to new settings if trained on an appropriately diverse set of environment and task configurations. Unsupervised Environment Design (UED) is a promising self-supervised RL paradigm, wherein the free parameters of an underspecified environment are automatically adapted during training to the agent's capabilities, leading to the emergence of diverse training environments. Here, we cast Prioritized Level Replay (PLR), an empirically successful but theoretically unmotivated method that selectively samples randomly-generated training levels, as UED. We argue that by curating completely random levels, PLR, too, can generate novel and complex levels for effective training. This insight reveals a natural class of UED methods we call Dual Curriculum Design (DCD). Crucially, DCD includes both PLR and a popular UED algorithm, PAIRED, as special cases and inherits similar theoretical guarantees. This connection allows us to develop novel theory for PLR, providing a version with a robustness guarantee at Nash equilibria. Furthermore, our theory suggests a highly counterintuitive improvement to PLR: by stopping the agent from updating its policy on uncurated levels (training on less data), we can improve the convergence to Nash equilibria. Indeed, our experiments confirm that our new method, PLR$^{\perp}$, obtains better results on a suite of out-of-distribution, zero-shot transfer tasks, in addition to demonstrating that PLR$^{\perp}$ improves the performance of PAIRED, from which it inherited its theoretical framework.

Institutions are increasingly relying on machine learning models Given the imbalanced nature of data in this domain, which makes to identify and alert on abnormal events, such as fraud, cyber attacks learning classifiers that efficiently discriminate among the minority and system failures. These alerts often need to be manually and majority class difficult, and the limited resources available investigated by specialists. Given the operational cost of manual inspections, for inspecting time-sensitive risky events, we are interested in understanding the suspicious events are selected by alerting systems with the relationship between the rate of detection from the carefully designed thresholds. In this paper, we consider an imbalanced minority class (i.e., the fraction of samples from the minority class binary classification problem, where events arrive sequentially selected for inspection) and the inspection budget. Specifically, we and only a limited number of suspicious events can be inspected. We focus on applications that involve real-time processing and decisionmaking model the event arrivals as a non-homogeneous Poisson process, and where an abnormal event can only be inspected at the time compare various suspicious event selection methods including those of arrival, and we investigate how different selection policies based based on static and adaptive thresholds. For each method, we analytically on classifier predictions operate in terms of the limited inspection characterize the tradeoff between the minority-class detection budget rather than the decision threshold.

With the technological advances in machine learning, effective ways are available to process the huge amount of data generated in real life. However, issues of privacy and scalability will constrain the development of machine learning. Federated learning (FL) can prevent privacy leakage by assigning training tasks to multiple clients, thus separating the central server from the local devices. However, FL still suffers from shortcomings such as single-point-failure and malicious data. The emergence of blockchain provides a secure and efficient solution for the deployment of FL. In this paper, we conduct a comprehensive survey of the literature on blockchained FL (BCFL). First, we investigate how blockchain can be applied to federal learning from the perspective of system composition. Then, we analyze the concrete functions of BCFL from the perspective of mechanism design and illustrate what problems blockchain addresses specifically for FL. We also survey the applications of BCFL in reality. Finally, we discuss some challenges and future research directions.

We characterize optimal oversight of algorithms in a world where an agent designs a complex prediction function but a principal is limited in the amount of information she can learn about the prediction function. We show that limiting agents to prediction functions that are simple enough to be fully transparent is inefficient as long as the bias induced by misalignment between principal's and agent's preferences is small relative to the uncertainty about the true state of the world. Algorithmic audits can improve welfare, but the gains depend on the design of the audit tools. Tools that focus on minimizing overall information loss, the focus of many post-hoc explainer tools, will generally be inefficient since they focus on explaining the average behavior of the prediction function rather than sources of mis-prediction, which matter for welfare-relevant outcomes. Targeted tools that focus on the source of incentive misalignment, e.g., excess false positives or racial disparities, can provide first-best solutions. We provide empirical support for our theoretical findings using an application in consumer lending.

At least since the 1950s, the idea that it would be possible to soon create a machine that was capable of matching the full scope and level of achievement of human intelligence has been greeted with equal amounts of hype and hysteria. We've now succeeded in creating machines that can solve specific fairly narrow problems -- "smart" machines that can diagnose disease, drive cars, understand speech, and beat us at chess -- but general intelligence remains elusive. Let's get this out of the way: Improvements in machine intelligence will not lead to runaway machine-led revolutions. They may change the kind of jobs that people do, but they will not spell the end of human existence. There will be no robo-apocalypse. The emphasis of intelligence testing and computational approaches to intelligence has been on well-structured and formal problems. That is, problems that have a clear goal and a set number of possible solutions. But we humans are creative, irrational, and inconsistent.

Amazon and MIT today announced the establishment of the Science Hub, a collaboration that will focus on areas of mutual interest, beginning with artificial intelligence and robotics in the first year. To get the hub started, Amazon will provide gift and sponsored research funding over the next five years to support research and academic fellowships on campus. The primary goals of the hub are to ensure the benefits of AI and robotics innovations are shared broadly -- both through education and by advancing research -- and to broaden participation in the research from diverse, interdisciplinary scholars, and other innovators. "AI and robotics have an enormous impact on every aspect of our lives, fundamentally changing how we work, learn, access resources and services, and connect to one another -- so it's critical we conduct research that advances the field in ways that are responsible, effective, and beneficial to society," said Aude Oliva, a senior research scientist and director of strategic industry engagement in the MIT Schwarzman College of Computing. "We take an expansive view of AI and robotics to include expertise from across all five of the Institute's schools. We're excited by the potential of collaborations with industry leaders who bring their insights to the research, want to support the next generation of talent, and are best positioned to implement what is learned."

The emergence of data science as a field of study and practical application over the last century has led to the development of technologies such as deep learning, natural language processing, and computer vision. Broadly speaking, it has enabled the emergence of machine learning (ML) as a way of working towards what we refer to as artificial intelligence (AI), a field of technology that's rapidly transforming the way we work and live. Data science encompasses the theoretical and practical application of ideas, including Big Data, predictive analytics, and artificial intelligence. If data is the oil of the information age and ML is the engine, then data science is the digital domain's equivalent of the laws of physics that cause combustion to occur and pistons to move. A key point to remember is that as the importance of understanding how to work with data grows, the science behind it is becoming more accessible.

It can be difficult to assess the quality of a fitted model when facing unsupervised learning problems. Latent variable models, such as variation autoencoders and Gaussian mixture models, are often trained with likelihood-based approaches. In scope of Goodhart's law, when a metric becomes a target it ceases to be a good metric and therefore we should not use likelihood to assess the quality of the fit of these models. The solution we propose is a new metric for model comparison or regularization that relies on moments. The concept is to study the difference between the data moments and the model moments using a matrix norm, such as the Frobenius norm. We show how to use this new metric for model comparison and then for regularization. It is common to draw samples from the fitted distribution when evaluating latent variable models and we show that our proposed metric is faster to compute and has a smaller variance that this alternative. We conclude this article with a proof of concept of both applications and we discuss future work.

Understanding human language has been a sub-challenge on the way of intelligent machines. The study of meaning in natural language processing (NLP) relies on the distributional hypothesis where language elements get meaning from the words that co-occur within contexts. The revolutionary idea of distributed representation for a concept is close to the working of a human mind in that the meaning of a word is spread across several neurons, and a loss of activation will only slightly affect the memory retrieval process. Neural word embeddings transformed the whole field of NLP by introducing substantial improvements in all NLP tasks. In this survey, we provide a comprehensive literature review on neural word embeddings. We give theoretical foundations and describe existing work by an interplay between word embeddings and language modelling. We provide broad coverage on neural word embeddings, including early word embeddings, embeddings targeting specific semantic relations, sense embeddings, morpheme embeddings, and finally, contextual representations. Finally, we describe benchmark datasets in word embeddings' performance evaluation and downstream tasks along with the performance results of/due to word embeddings.

An Intelligent IoT Environment (iIoTe) is comprised of heterogeneous devices that can collaboratively execute semi-autonomous IoT applications, examples of which include highly automated manufacturing cells or autonomously interacting harvesting machines. Energy efficiency is key in such edge environments, since they are often based on an infrastructure that consists of wireless and battery-run devices, e.g., e-tractors, drones, Automated Guided Vehicle (AGV)s and robots. The total energy consumption draws contributions from multiple iIoTe technologies that enable edge computing and communication, distributed learning, as well as distributed ledgers and smart contracts. This paper provides a state-of-the-art overview of these technologies and illustrates their functionality and performance, with special attention to the tradeoff among resources, latency, privacy and energy consumption. Finally, the paper provides a vision for integrating these enabling technologies in ...