Imitation learning seeks to learn an expert policy from sampled demonstrations. However, in the real world, it is often difficult to find a perfect expert and avoiding dangerous behaviors becomes relevant for safety reasons. We present the idea of \textit{learning to avoid}, an objective opposite to imitation learning in some sense, where an agent learns to avoid a demonstrator policy given an environment. We define avoidance learning as the process of optimizing the agent's reward while avoiding dangerous behaviors given by a demonstrator. In this work we develop a framework of avoidance learning by defining a suitable objective function for these problems which involves the \emph{distance} of state occupancy distributions of the expert and demonstrator policies. We use density estimates for state occupancy measures and use the aforementioned distance as the reward bonus for avoiding the demonstrator. We validate our theory with experiments using a wide range of partially observable environments. Experimental results show that we are able to improve sample efficiency during training compared to state of the art policy optimization and safety methods.

In this paper, we study the convergence of a class of gradient-based Model-Agnostic Meta-Learning (MAML) methods and characterize their overall computational complexity as well as their best achievable level of accuracy in terms of gradient norm for nonconvex loss functions. In particular, we start with the MAML algorithm and its first order approximation (FO-MAML) and highlight the challenges that emerge in their analysis. By overcoming these challenges not only we provide the first theoretical guarantees for MAML and FO-MAML in nonconvex settings, but also we answer some of the unanswered questions for the implementation of these algorithms including how to choose their learning rate (stepsize) and the batch size for both tasks and datasets corresponding to tasks. In particular, we show that MAML can find an $\epsilon$-first-order stationary point for any positive $\epsilon$ after at most $\mathcal{O}(1/\epsilon^2)$ iterations at the expense of requiring second-order information. We also show that the FO-MAML method which ignores the second-order information required in the update of MAML cannot achieve any small desired level of accuracy, i.e, FO-MAML cannot find an $\epsilon$-first-order stationary point for any positive $\epsilon$. We further propose a new variant of the MAML algorithm called Hessian-free MAML (HF-MAML) which preserves all theoretical guarantees of MAML, without requiring access to the second-order information of loss functions.

Immersive simulations are increasingly used for teaching and training in many societally important arenas including healthcare, disaster response and science education. The interactions of participants in such settings lead to a complex array of emergent outcomes that present challenges for analysis. This paper studies a central element of such an analysis, namely the interpretability of models for inferring structure in time series data. This problem is explored in the context of modeling student interactions in an immersive ecological-system simulation. Unsupervised machine learning is applied to data on system dynamics with the aim of helping teachers determine the effects of students' actions on these dynamics. We address the question of choosing the optimal machine learning model, considering both statistical information criteria and interpretabilty quality. The results of a user study show that the models that are the best understood by people are not those that optimize information theoretic criteria. In addition, a model using a fully Bayesian approach performed well on both statistical measures and on human-subject tests of interpretabilty, making it a good candidate for automated model selection that does not require human-in-the-loop evaluation. The results from this paper are already being used in the classroom and can inform the design of interpretable models for a broad range of socially relevant domains. 1 Introduction There is increasing evidence of the value of multi-person embodied simulations for engaging learners in a variety of applications, such as healthcare, disaster response and education (Alinier et al. 2014; Amir and Gal 2013).

We study the problem of learning properties of nodes in tree structures. Those properties are specified by logical formulas, such as formulas from first-order or monadic second-order logic. We think of the tree as a database encoding a large dataset and therefore aim for learning algorithms which depend at most sublinearly on the size of the tree. We present a learning algorithm for quantifier-free formulas where the running time only depends polynomially on the number of training examples, but not on the size of the background structure. By a previous result on strings we know that for general first-order or monadic second-order (MSO) formulas a sublinear running time cannot be achieved. However, we show that by building an index on the tree in a linear time preprocessing phase, we can achieve a learning algorithm for MSO formulas with a logarithmic learning phase.

YOLO is a social robot designed and developed to stimulate creativity in children through storytelling activities. Children use it as a character in their stories. This article details the artificial intelligence software developed for YOLO. The implemented software schedules through several Creativity Behaviors to find the ones that stimulate creativity more effectively. YOLO can choose between convergent and divergent thinking techniques, two important processes of creative thought. These techniques were developed based on the psychological theories of creativity development and on research from creativity experts who work with children. Additionally, this software allows the creation of Social Behaviors that enable the robot to behave as a believable character. On top of our framework, we built 3 main social behavior parameters: Exuberant, Aloof, and Harmonious. These behaviors are meant to ease immersive play and the process of character creation. The 3 social behaviors were based on psychological theories of personality and developed using children's input during co-design studies. Overall, this work presents an attempt to design, develop, and deploy social robots that nurture intrinsic human abilities, such as the ability to be creative.

Knowledge Distillation (KD) aims to transfer knowledge in a teacher-student framework, by providing the predictions of the teacher network to the student network in the training stage to help the student network generalize better. It can use either a teacher with high capacity or {an} ensemble of multiple teachers. However, the latter is not convenient when one wants to use feature-map-based distillation methods. For a solution, this paper proposes a versatile and powerful training algorithm named FEature-level Ensemble for knowledge Distillation (FEED), which aims to transfer the ensemble knowledge using multiple teacher networks. We introduce a couple of training algorithms that transfer ensemble knowledge to the student at the feature map level. Among the feature-map-based distillation methods, using several non-linear transformations in parallel for transferring the knowledge of the multiple teacher{s} helps the student find more generalized solutions. We name this method as parallel FEED, andexperimental results on CIFAR-100 and ImageNet show that our method has clear performance enhancements, without introducing any additional parameters or computations at test time. We also show the experimental results of sequentially feeding teacher's information to the student, hence the name sequential FEED, and discuss the lessons obtained. Additionally, the empirical results on measuring the reconstruction errors at the feature map give hints for the enhancements.

Language model pre-training, such as BERT, has significantly improved the performances of many natural language processing tasks. However, pre-trained language models are usually computationally expensive and memory intensive, so it is difficult to effectively execute them on some resource-restricted devices. To accelerate inference and reduce model size while maintaining accuracy, we firstly propose a novel transformer distillation method that is a specially designed knowledge distillation (KD) method for transformer-based models. By leveraging this new KD method, the plenty of knowledge encoded in a large teacher BERT can be well transferred to a small student TinyBERT. Moreover, we introduce a new two-stage learning framework for TinyBERT, which performs transformer distillation at both the pre-training and task-specific learning stages. This framework ensures that TinyBERT can capture both the general-domain and task-specific knowledge of the teacher BERT. TinyBERT is empirically effective and achieves comparable results with BERT in GLUE datasets, while being 7.5x smaller and 9.4x faster on inference. TinyBERT is also significantly better than state-of-the-art baselines, even with only about 28% parameters and 31% inference time of baselines.

Despite its role in early 20th-century fiction, AI has been part of the professional conversation for barely 70 years. AI was first studied at a Dartmouth College conference in 1956. The 1960's saw gains in machine translation and analysis. But AI underwent a "winter" from the 1970s through the early '90s. Researchers shelved their work largely because of the problem of "combinatorial explosion." A U.K. professor who first described the AI concept worried that too many variables would make it useless outside of lab settings. In the early '70s, groups like the U.S. Defense Advanced Research Projects Agency pulled funding. Research failures had become the norm. Interest in AI grew during the 1990s and early 2000's. Processing power and data volumes increased. At the same time, data sets grew massively. Algorithms gained more "meat" on which to train. Advances in game theory and data modeling led to new approaches. Today, best-in-class infrastructures can support 100,000 or more computers. Two-and-a-half quintillion bytes of data are now generated every day. Globally, private firms are spending tens of billions of dollars per year researching and improving AI initiatives. In fact, 2018's investment amount is more than 50 percent larger than last year's alone. Add it all up, and AI seems ready for a leap forward unlike any seen in its history. But, after slow decades followed by speedy discoveries, few outside the field feel they truly understand it. A recent Dell Technologies report found that 67 percent of leaders said their companies were struggling to implement AI. A similar two out of three consumers don't even realize they're using it, according to a HubSpot survey. "By far the greatest danger of artificial intelligence is that people conclude too early that they understand it."

Having first appeared on the scene of machine learning in 1986 and artificial neural networks in 2000, the study of deep learning continues to explode with new research, advanced techniques, higher benchmarks, and broader applications. Keeping pace in such an active field with an average of 30 new deep learning papers uploaded to arXiv per day over the previous month is daunting, to say the least. While there are many key deep learning scientists and engineers active today, the following list of 12 researchers and innovators in the field are among the most important – and they so happen to actively share on social media, making their progress and insights much easier to keep up with. So, start paying attention to these 12 top deep learning individuals, and be prepared to expand your understanding and awareness of the incredible advancements deep learning is bringing to science, industry, and society. While it in no way correlates to everyone's contribution to the field, the list is sorted by the number of Twitter followers so you can see who appears to have the most reach today.

Recently, a study indicated UK workers fearing artificial intelligence to eliminate their jobs. Is AI going to steal away our jobs or is it all a hype? We're undergoing a technology revolution experiencing breathless stories about how artificial intelligence is going to replace our jobs. Are these signs of artificial intelligence apocalypse? Is there anything at all that reveals the AI phenomenon or is it just an image of a boogeyman trying to scare us into line?