The notion of uncertainty is of major importance in machine learning and constitutes a key element of machine learning methodology. In line with the statistical tradition, uncertainty has long been perceived as almost synonymous with standard probability and probabilistic predictions. Yet, due to the steadily increasing relevance of machine learning for practical applications and related issues such as safety requirements, new problems and challenges have recently been identified by machine learning scholars, and these problems may call for new methodological developments. In particular, this includes the importance of distinguishing between (at least) two different types of uncertainty, often refereed to as aleatoric and epistemic. In this paper, we provide an introduction to the topic of uncertainty in machine learning as well as an overview of hitherto attempts at handling uncertainty in general and formalizing this distinction in particular. 1 Introduction Machine learning is essentially concerned with extracting models from data and using these models to make predictions.

We present local ensembles, a method for detecting extrapolation at test time in a pre-trained model. We focus on underdetermination as a key component of extrapolation: we aim to detect when many possible predictions are consistent with the training data and model class. Our method uses local second-order information to approximate the variance of predictions across an ensemble of models from the same class. We compute this approximation by estimating the norm of the component of a test point's gradient that aligns with the low-curvature directions of the Hessian, and provide a tractable method for estimating this quantity. Experimentally, we show that our method is capable of detecting when a pre-trained model is extrapolating on test data, with applications to out-of-distribution detection, detecting spurious correlates, and active learning.

Deep reinforcement learning (DRL) is a booming area of artificial intelligence. Many practical applications of DRL naturally involve more than one collaborative learners, making it important to study DRL in a multi-agent context. Previous research showed that effective learning in complex multi-agent systems demands for highly coordinated environment exploration among all the participating agents. Many researchers attempted to cope with this challenge through learning centralized value functions. However, the common strategy for every agent to learn their local policies directly often fail to nurture strong inter-agent collaboration and can be sample inefficient whenever agents alter their communication channels. To address these issues, we propose a new framework known as centralized training and exploration with decentralized execution via policy distillation. Guided by this framework and the maximum-entropy learning technique, we will first train agents' policies with shared global component to foster coordinated and effective learning. Locally executable policies will be derived subsequently from the trained global policies via policy distillation. Experiments show that our new framework and algorithm can achieve significantly better performance and higher sample efficiency than a cutting-edge baseline on several multi-agent DRL benchmarks.

Dynamic games are an effective paradigm for dealing with the control of multiple interacting actors. Current algorithms for solving these problems either rely on Hamilton-Jacobi-Isaacs (HJI) methods, dynamic programming (DP), differential dynamic programming (DDP), or an iterative best response approach (IBR). The first two approaches have strong theoretical guarantees; however they becomes intractable in high-dimensional real-world applications. The third approach is grounded in the success of iLQR. It is scalable, but it cannot handle constraints. Finally, the iterative best response algorithm is a heuristic approach with unknown convergence properties, and it can suffer from stability and tractability issues. This paper introduces ALGAMES (Augmented Lagrangian GAME-theoretic Solver), a solver that handles trajectory optimization problems with multiple actors and general nonlinear state and input constraints. We evaluate our solver in the context of autonomous driving on scenarios involving numerous vehicles such as ramp merging, overtaking, and lane changing. We present simulation and timing results demonstrating the speed and the ability of the solver to produce efficient, safe, and natural autonomous behaviors.

Entity Coreference Resolution is the task of resolving all the mentions in a document that refer to the same real world entity and is considered as one of the most difficult tasks in natural language understanding. While in it is not an end task, it has been proved to improve downstream natural language processing tasks such as entity linking, machine translation, summarization and chatbots. We conducted a systematic a review of neural-based approached and provide a detailed appraisal of the datasets and evaluation metrics in the field. Emphasis is given on Pronoun Resolution, a subtask of Coreference Resolution, which has seen various improvements in the recent years. We conclude the study by highlight the lack of agreed upon standards and propose a way to expand the task even further.

Artificial intelligence (AI) expert and Flamingo Ai Founder and Executive Director Dr Catriona Wallace is set to share her insights on what we can look forward to in a world with more advanced AI, at this year's CEBIT expo. The keynote, titled'AI: Human Machine: Who gets the upper hand?' will explore developments in AI, how it's being used and how it will transform the business world and life as we know it. "AI, described as the most powerful force equal in impact to the discovery of fire and the invention of electricity, will increasingly become the primary power driving the massive changes that [climate change and disruptive technologies] will bring," Wallace said. "With AI set to replace 40% of jobs and 30% of business interactions in the next five years, and the time of'singularity', where machines may become'smarter' than humans possibly just 20 years away, the onus will be on people to successfully navigate the Fourth Industrial Revolution." NSW Minister for Jobs and Investment Stuart Ayres said CEBIT Australia will provide an international forum for technology companies to do business and discuss the future, including the impact of AI and how it can be harnessed to secure new jobs.

Alphabet (Google) subsidiary Wing has become the first company in the United States to deliver packages by drone. In Christiansburg, the small Virginia town chosen as Wing's test location, the 22,000 residents can order products normally shipped by FedEx, medicine from Walgreens and a selection of candy from a local business -- all of which will arrive via drone. Wing, which already operates in two Australian cities as well as Helsinki, announced in a statement that the first drone-powered deliveries had taken place Friday afternoon in Christiansburg, "paving the way for the most advanced drone delivery service in the nation". One family used the Wing app to order Tylenol, cough drops, Vitamin C tablets, bottled water and tissues, the statement said. An older resident ordered a birthday present for his wife.

I would like a large cheese pizza with an ominous side of surveillance, please. Earlier this year, Domino's, the worldwide purveyor of mediocre pizza, introduced a snazzy tool called the Dom Pizza Checker to its Australia and New Zealand locations. According to its website, in-store cameras "use advanced machine learning, artificial intelligence and sensor technology to identify pizza type, even topping distribution and correct toppings". If your food doesn't match your order, or internal quality standards, workers are ordered to make it again. Basically, Big Brother is watching your pizza.

We live in a connected world and generate a vast amount of connected data. Social networks, financial transaction systems, biological networks, transportation systems and a telecommunication nexus are all examples. The paper citation network displayed in Figure 1 is another example of connected data. Representing connected data is possible using a graph data structure regularly used in Computer Science. In this article, we will provide an introduction to the assorted types of connected data, what they represent, and the challenges we can solve.

Researchers from China and South Korea have created an AI that can predict El Niño up to 18 months before it occurs. El Niño is a weather event that can occur every 2-7 years, where the area of warmer water in the western Pacific Ocean around Australia spreads across the Pacific. This leads to warmer air rising across the Pacific, causing severe rainfall and drastically changing wind direction and strength across the Pacific. This has huge knock on effects on weather worldwide. El Niño can cause colder winters in northern Europe and droughts in countries such as Australia and Malaysia.