Contextual bandits serve as a fundamental model for many sequential decision making tasks. The most popular theoretically justified approaches are based on the optimism principle. While these algorithms can be practical, they are known to be suboptimal asymptotically (Lattimore and Szepesvari, 2017). On the other hand, existing asymptotically optimal algorithms for this problem do not exploit the linear structure in an optimal way and suffer from lower-order terms that dominate the regret in all practically interesting regimes. We start to bridge the gap by designing an algorithm that is asymptotically optimal and has good finite-time empirical performance. At the same time, we make connections to the recent literature on when exploration-free methods are effective. Indeed, if the distribution of contexts is well behaved, then our algorithm acts mostly greedily and enjoys sub-logarithmic regret. Furthermore, our approach is adaptive in the sense that it automatically detects the nice case. Numerical results demonstrate significant regret reductions by our method relative to several baselines.

We present a modern scalable reinforcement learning agent called SEED (Scalable, Efficient Deep-RL). By effectively utilizing modern accelerators, we show that it is not only possible to train on millions of frames per second but also to lower the cost of experiments compared to current methods. We achieve this with a simple architecture that features centralized inference and an optimized communication layer. SEED adopts two state of the art distributed algorithms, IMPALA/V-trace (policy gradients) and R2D2 (Q-learning), and is evaluated on Atari-57, DeepMind Lab and Google Research Football. We improve the state of the art on Football and are able to reach state of the art on Atari-57 twice as fast in wall-time. For the scenarios we consider, a 40% to 80% cost reduction for running experiments is achieved. The implementation along with experiments is open-sourced so that results can be reproduced and novel ideas tried out.

Model-free reinforcement learning is known to be memory and computation efficient and more amendable to large scale problems. In this paper, two model-free algorithms are introduced for learning infinite-horizon average-reward Markov Decision Processes (MDPs). The first algorithm reduces the problem to the discounted-reward version and achieves $\mathcal{O}(T^{2/3})$ regret after $T$ steps, under the minimal assumption of weakly communicating MDPs. The second algorithm makes use of recent advances in adaptive algorithms for adversarial multi-armed bandits and improves the regret to $\mathcal{O}(\sqrt{T})$, albeit with a stronger ergodic assumption. To the best of our knowledge, these are the first model-free algorithms with sub-linear regret (that is polynomial in all parameters) in the infinite-horizon average-reward setting.

Aerial cinematography is revolutionizing industries that require live and dynamic camera viewpoints such as entertainment, sports, and security. However, safely piloting a drone while filming a moving target in the presence of obstacles is immensely taxing, often requiring multiple expert human operators. Hence, there is demand for an autonomous cinematographer that can reason about both geometry and scene context in real-time. Existing approaches do not address all aspects of this problem; they either require high-precision motion-capture systems or GPS tags to localize targets, rely on prior maps of the environment, plan for short time horizons, or only follow artistic guidelines specified before flight. In this work, we address the problem in its entirety and propose a complete system for real-time aerial cinematography that for the first time combines: (1) vision-based target estimation; (2) 3D signed-distance mapping for occlusion estimation; (3) efficient trajectory optimization for long time-horizon camera motion; and (4) learning-based artistic shot selection. We extensively evaluate our system both in simulation and in field experiments by filming dynamic targets moving through unstructured environments. Our results indicate that our system can operate reliably in the real world without restrictive assumptions. We also provide in-depth analysis and discussions for each module, with the hope that our design tradeoffs can generalize to other related applications. Videos of the complete system can be found at: https://youtu.be/ookhHnqmlaU.

According to a mainstream position in contemporary cognitive science and philosophy, the use of abstract compositional concepts is both a necessary and a sufficient condition for the presence of genuine thought. In this article, we show how the ability to develop and utilise abstract conceptual structures can be achieved by a particular kind of learning agents. More specifically, we provide and motivate a concrete operational definition of what it means for these agents to be in possession of abstract concepts, before presenting an explicit example of a minimal architecture that supports this capability. We then proceed to demonstrate how the existence of abstract conceptual structures can be operationally useful in the process of employing previously acquired knowledge in the face of new experiences, thereby vindicating the natural conjecture that the cognitive functions of abstraction and generalisation are closely related. Keywords: concept formation, projective simulation, reinforcement learning, transparent artificial intelligence, theory formation, explainable artificial intelligence (XAI)

An AI learns to park a car in a parking lot in a 3D physics simulation. The AI consists of a deep Neural Network with 3 hidden layers of 128 neurons each. It is trained with the Proximal Policy Optimization (PPO) algorithm, which is a Reinforcement Learning approach. Basically, the input of the Neural Network are the readings of eight depth sensors, the cars current speed and position, as well as its relative position to the target. The outputs of the Neural Network are interpreted as engine force, braking force and turning force.

Drawing on recent advances in Deep Imitation Learning and Deep Reinforcement Learning, covariant.ai is developing AI software that makes it easy to teach robots new, complex skills. Founded by Pieter Abbeel, Peter Chen, Rocky Duan and Tianhao Zhang, the company is based in Emeryville, CA and backed by venture funding.

Drawing on recent advances in Deep Imitation Learning and Deep Reinforcement Learning, covariant.ai is developing AI software that makes it easy to teach robots new, complex skills. Founded by Pieter Abbeel, Peter Chen, Rocky Duan and Tianhao Zhang, the company is based in Emeryville, CA and backed by venture funding.

This paper considers a semi-supervised learning framework for weakly labeled polyphonic sound event detection problems for the DCASE 2019 challenge's task4 by combining both the tri-training and adversarial learning. The goal of the task4 is to detect onsets and offsets of multiple sound events in a single audio clip. The entire dataset consists of the synthetic data with a strong label (sound event labels with boundaries) and real data with weakly labeled (sound event labels) and unlabeled dataset. Given this dataset, we apply the tri-training where two different classifiers are used to obtain pseudo labels on the weakly labeled and unlabeled dataset, and the final classifier is trained using the strongly labeled dataset and weakly/unlabeled dataset with pseudo labels. Also, we apply the adversarial learning to reduce the domain gap between the real and synthetic dataset. We evaluated our learning framework using the validation set of the task4 dataset, and in the experiments, our learning framework shows a considerable performance improvement over the baseline model.

Neuroscientific theory suggests that dopaminergic neurons broadcast global reward prediction errors to large areas of the brain influencing the synaptic plasticity of the neurons in those regions. We build on this theory to propose a multi-agent learning framework with spiking neurons in the generalized linear model (GLM) formulation as agents, to solve reinforcement learning (RL) tasks. We show that a network of GLM spiking agents connected in a hierarchical fashion, where each spiking agent modulates its firing policy based on local information and a global prediction error, can learn complex action representations to solve RL tasks. We further show how leveraging principles of modularity and population coding inspired from the brain can help reduce variance in the learning updates making it a viable optimization technique.