In reinforcement learning (RL), when defining a Markov Decision Process (MDP), the environment dynamics is implicitly assumed to be stationary. This assumption of stationarity, while simplifying, can be unrealistic in many scenarios. In the continual reinforcement learning scenario, the sequence of tasks is another source of nonstationarity. In this work, we propose to examine this continual reinforcement learning setting through the block contextual MDP (BC-MDP) framework, which enables us to relax the assumption of stationarity. This framework challenges RL algorithms to handle both nonstationarity and rich observation settings and, by additionally leveraging smoothness properties, enables us to study generalization bounds for this setting. Finally, we take inspiration from adaptive control to propose a novel algorithm that addresses the challenges introduced by this more realistic BC-MDP setting, allows for zero-shot adaptation at evaluation time, and achieves strong performance on several nonstationary environments.

We consider an extended notion of reinforcement learning in which the environment can simulate the agent and base its outputs on the agent's hypothetical behavior. Since good performance usually requires paying attention to whatever things the environment's outputs are based on, we argue that for an agent to achieve on-average good performance across many such extended environments, it is necessary for the agent to self-reflect. Thus, an agent's self-reflection ability can be numerically estimated by running the agent through a battery of extended environments. We are simultaneously releasing an open-source library of extended environments to serve as proof-of-concept of this technique. As the library is first-of-kind, we have avoided the difficult problem of optimizing it. Instead we have chosen environments with interesting properties. Some seem paradoxical, some lead to interesting thought experiments, some are even suggestive of how self-reflection might have evolved in nature. We give examples and introduce a simple transformation which experimentally seems to increase self-reflection.

When learning common skills like driving, beginners usually have domain experts standing by to ensure the safety of the learning process. We formulate such learning scheme under the Expert-in-the-loop Reinforcement Learning where a guardian is introduced to safeguard the exploration of the learning agent. While allowing the sufficient exploration in the uncertain environment, the guardian intervenes under dangerous situations and demonstrates the correct actions to avoid potential accidents. Thus ERL enables both exploration and expert's partial demonstration as two training sources. Following such a setting, we develop a novel Expert Guided Policy Optimization (EGPO) method which integrates the guardian in the loop of reinforcement learning. The guardian is composed of an expert policy to generate demonstration and a switch function to decide when to intervene. Particularly, a constrained optimization technique is used to tackle the trivial solution that the agent deliberately behaves dangerously to deceive the expert into taking over. Offline RL technique is further used to learn from the partial demonstration generated by the expert. Safe driving experiments show that our method achieves superior training and test-time safety, outperforms baselines with a substantial margin in sample efficiency, and preserves the generalizabiliy to unseen environments in test-time. Demo video and source code are available at: https://decisionforce.github.io/EGPO/

We study multilingual AMR parsing from the perspective of knowledge distillation, where the aim is to learn and improve a multilingual AMR parser by using an existing English parser as its teacher. We constrain our exploration in a strict multilingual setting: there is but one model to parse all different languages including English. We identify that noisy input and precise output are the key to successful distillation. Together with extensive pre-training, we obtain an AMR parser whose performances surpass all previously published results on four different foreign languages, including German, Spanish, Italian, and Chinese, by large margins (up to 18.8 \textsc{Smatch} points on Chinese and on average 11.3 \textsc{Smatch} points). Our parser also achieves comparable performance on English to the latest state-of-the-art English-only parser.

Bilevel optimization (BO) has arisen as a powerful tool for solving many modern machine learning problems. However, due to the nested structure of BO, existing gradient-based methods require second-order derivative approximations via Jacobian- or/and Hessian-vector computations, which can be very costly in practice, especially with large neural network models. In this work, we propose a novel BO algorithm, which adopts Evolution Strategies (ES) based method to approximate the response Jacobian matrix in the hypergradient of BO, and hence fully eliminates all second-order computations. We call our algorithm as ESJ (which stands for the ES-based Jacobian method) and further extend it to the stochastic setting as ESJ-S. Theoretically, we characterize the convergence guarantee and computational complexity for our algorithms. Experimentally, we demonstrate the superiority of our proposed algorithms compared to the state of the art methods on various bilevel problems. Particularly, in our experiment in the few-shot meta-learning problem, we meta-learn the twelve millions parameters of a ResNet-12 network over the miniImageNet dataset, which evidently demonstrates the scalability of our ES-based bilevel approach and its feasibility in the large-scale setting.

Belonging to the family of Bayesian nonparametrics, Gaussian process (GP) based approaches have well-documented merits not only in learning over a rich class of nonlinear functions, but also in quantifying the associated uncertainty. However, most GP methods rely on a single preselected kernel function, which may fall short in characterizing data samples that arrive sequentially in time-critical applications. To enable {\it online} kernel adaptation, the present work advocates an incremental ensemble (IE-) GP framework, where an EGP meta-learner employs an {\it ensemble} of GP learners, each having a unique kernel belonging to a prescribed kernel dictionary. With each GP expert leveraging the random feature-based approximation to perform online prediction and model update with {\it scalability}, the EGP meta-learner capitalizes on data-adaptive weights to synthesize the per-expert predictions. Further, the novel IE-GP is generalized to accommodate time-varying functions by modeling structured dynamics at the EGP meta-learner and within each GP learner. To benchmark the performance of IE-GP and its dynamic variant in the adversarial setting where the modeling assumptions are violated, rigorous performance analysis has been conducted via the notion of regret, as the norm in online convex optimization. Last but not the least, online unsupervised learning for dimensionality reduction is explored under the novel IE-GP framework. Synthetic and real data tests demonstrate the effectiveness of the proposed schemes.

The transition towards data-centric AI requires revisiting data notions from mathematical and implementational standpoints to obtain unified data-centric machine learning packages. Towards this end, this work proposes unifying principles offered by categorical and cochain notions of data, and discusses the importance of these principles in data-centric AI transition. In the categorical notion, data is viewed as a mathematical structure that we act upon via morphisms to preserve this structure. As for cochain notion, data can be viewed as a function defined in a discrete domain of interest and acted upon via operators. While these notions are almost orthogonal, they provide a unifying definition to view data, ultimately impacting the way machine learning packages are developed, implemented, and utilized by practitioners.

Thousands of working adults will soon benefit from free courses that will help them to rapidly upskill or retrain, as part of the government's drive to plug skills gaps and boost access to more high-quality training alternatives. Sixty-five short and modular courses will start to roll out from later this month at 10 Institutes of Technology (IoTs) across England in sought-after STEM subjects. This will include courses such as Artificial Intelligence, Digitisation of Manufacturing, Digital Construction, Agricultural Robotics, and Cyber Security. The courses will be a blend of classroom and remote online study, and will vary in length from 50 to 138 hours – giving more adults greater flexibility in how and when they learn, so they can fit it around their lives. Swindon and Wiltshire IoT, for instance, will offer five short 50 hour courses across eight weeks.

In our mission towards powering careers through tech education, we are doubling down on the support that we offer to our students. As a part of this, we want to extend an opportunity to you to become a Mentor in our scholarship programs. Career mentors are industry professionals who provide world-class career support to our students in their Nanodegree journey or beyond. All our mentors are handpicked through a rigorous selection process so that we can truly be #studentfirst. There are only a few positions, so please don't wait too long to apply!

When you create machine learning models in the real world, as opposed to online courses or Kaggle style competitions, you need to generate the training data yourself. This is the first step in any ML project. And as crucial as it is, it is the one we overlook the most. As a freelance ML engineer, I take on new projects often and face the same problem again and again: How can I generate the training data, faster and better? In this article, I want to share some of the best practices I have discovered on the way and help you boost your productivity.