With the rapid pace of innovation continually disrupting business models, and in many cases entire industries, how will online learning keep up to provide the relevant courseware for today's and tomorrow's workforce? This will be essential for economic growth and to support a thriving, college-educated workforce that's equipped with the very latest knowledge, ideas and technology. In the future, I believe that institutions at the forefront of online education will be recognized via several capabilities which will have digitally transformed today's EdTech market. They will include a powerful combination of omni-channel learning pathways, cognitive courseware, virtual counselors and AI-enabled course development and grading. These innovations, underpinned by artificial intelligence (AI), will help to provide students the ultimate choice in their courseware – including up-to-the-minute courses on high-interest/high-growth subject matter – as well as highly-innovative digital services that support them every step of the way to help maximize their success and personal objectives.

While the debate regarding how much screen time is appropriate for children rages on among educators, psychologists, and parents, it's another emerging technology in the form of artificial intelligence and machine learning that is beginning to alter education tools and institutions and changing what the future might look like in education. It is expected that artificial intelligence in U.S. education will grow by 47.5% from 2017-2021 according to the Artificial Intelligence Market in the US Education Sector report. Even though most experts believe the critical presence of teachers is irreplaceable, there will be many changes to a teacher's job and to educational best practices. AI has already been applied to education primarily in some tools that help develop skills and testing systems. As AI educational solutions continue to mature, the hope is that AI can help fill needs gaps in learning and teaching and allow schools and teachers to do more than ever before.

EARTHMOVER-BASED MANIFOLD LEARNING FOR ANAL YZING MOLECULAR CONFORMA TION SPACES Nathan Zelesko Amit Moscovich Joe Kileel Amit Singer, Department of Mathematics, Brown University Program in Applied and Computational Mathematics, Princeton University Department of Mathematics, Princeton University ABSTRACT In this paper, we propose a novel approach for manifold learning that combines the Earthmover's distance (EMD) with the diffusion maps method for dimensionality reduction. We demonstrate the potential benefits of this approach for learning shape spaces of proteins and other flexible macromolecules using a simulated dataset of 3-D density maps that mimic the nonuniform rotary motion of A TP synthase. Our results show that EMD-based diffusion maps require far fewer samples to recover the intrinsic geometry than the standard diffusion maps algorithm that is based on the Euclidean distance. To reduce the computational burden of calculating the EMD for all volume pairs, we employ a wavelet-based approximation to the EMD which reduces the computation of the pairwise EMDs to a computation of pairwise weighted-null 1 distances between wavelet coefficient vectors. Index T erms -- Shape space, dimensionality reduction, Wasserstein metric, diffusion maps, Laplacian eigenmaps, cryo-electron microscopy 1. INTRODUCTION Proteins and other macromolecules are elastic structures that may deform in various ways.

--Catastrophic forgetting continues to severely restrict the learnability of controllers suitable for multiple task environments. Efforts to combat catastrophic forgetting reported in the literature to date have focused on how control systems can be updated more rapidly, hastening their adjustment from good initial settings to new environments, or more circumspectly, suppressing their ability to overfit to any one environment. When using robots, the environment includes the robot's own body, its shape and material properties, and how its actuators and sensors are distributed along its mechanical structure. Here we demonstrate for the first time how one such design decision (sensor placement) can alter the landscape of the loss function itself, either expanding or shrinking the weight manifolds containing suitable controllers for each individual task, thus increasing or decreasing their probability of overlap across tasks, and thus reducing or inducing the potential for catastrophic forgetting. It has been shown in various single-task settings how an appropriate robot design can simplify the control problem [18, 27, 4, 2, 17, 22], but because these robots were restricted to a single training environment, they did not suffer catastrophic forgetting. Catastrophic forgetting is a major and unsolved challenge in the machine learning literature [9, 11, 15, 20].

Orthogonal Gradient Descent for Continual LearningMehrdad Farajtabar Navid Azizan 1 Alex Mott Ang Li DeepMind CalTech DeepMind DeepMind Abstract Neural networks are achieving state of the art and sometimes superhuman performance on learning tasks across a variety of domains. Whenever these problems require learning in a continual or sequential manner, however, neural networks suffer from the problem of catastrophic forgetting; they forget how to solve previous tasks after being trained on a new task, despite having the essential capacity to solve both tasks if they were trained on both simultaneously. In this paper, we propose to address this issue from a parameter space perspective and study an approach to restrict the direction of the gradient updates to avoid forgetting previously-learned data. We present the Orthogonal Gradient Descent (OGD) method, which accomplishes this goal by projecting the gradients from new tasks onto a subspace in which the neural network output on previous task does not change and the projected gradient is still in a useful direction for learning the new task. Our approach utilizes the high capacity of a neural network more efficiently and does not require storing the previously learned data that might raise privacy concerns. Experiments on common benchmarks reveal the effectiveness of the proposed OGD method. 1 Introduction One critical component of intelligence is the ability to learn continuously, when new information is constantly available but previously presented information is unavailable to retrieve. Despite their ubiquity in the real world, these problems have posed a longstanding challenge to artificial intelligence (Thrun and Mitchell, 1995; Hassabis et al., 2017).Correspondence to farajtabar@google.com. 1 Work done during an internship at DeepMind. A typical neural network training procedure over a sequence of different tasks usually results in degraded performance on previously trained tasks if the model could not revisit the data of previous tasks. This phenomenon is called catastrophic forgetting (McCloskey and Cohen, 1989; Ratcliff, 1990; French, 1999). Ideally, an intelligent agent should be able to learn consecutive tasks without degrading its performance on those already learned. With the deep learning renaissance (Krizhevsky et al., 2012; Hinton et al., 2006; Si-monyan and Zisserman, 2014) this problem has been revived (Srivastava et al., 2013; Goodfellow et al., 2013) with many followup studies (Parisi et al., 2019). One probable reason for this phenomenon is that neural networks are usually trained by Stochastic Gradient Descent (SGD)--or its variants--where the op-timizers produce gradients that are oblivious to past knowledge.

Variable metric proximal gradient (VM-PG) is a widely used class of convex optimization method. Lately, there has been a lot of research on the theoretical guarantees of VM-PG with different metric selections. However, most such metric selections are dependent on (an expensive) Hessian, or limited to scalar stepsizes like the Barzilai-Borwein (BB) stepsize with lots of safeguarding. Instead, in this paper we propose an adaptive metric selection strategy called the diagonal Barzilai-Borwein (BB) stepsize. The proposed diagonal selection better captures the local geometry of the problem while keeping per-step computation cost similar to the scalar BB stepsize i.e. $O(n)$. Under this metric selection for VM-PG, the theoretical convergence is analyzed. Our empirical studies illustrate the improved convergence results under the proposed diagonal BB stepsize, specifically for ill-conditioned machine learning problems for both synthetic and real-world datasets.

--Existing algorithms aiming to learn a binary classifier from positive (P) and unlabeled (U) data require estimating the class prior or label noise ahead of building a classification model. However, the estimation and classifier learning are normally conducted in a pipeline instead of being jointly optimized. In this paper, we propose to alternatively train the two steps using reinforcement learning. Our proposal adopts a policy network to adaptively make assumptions on the labels of unlabeled data, while a classifier is built upon the output of the policy network and provides rewards to learn a better policy. The dynamic and interactive training between the policy maker and the classifier can exploit the unlabeled data in a more effective manner and yield a significant improvement in terms of classification performance. Furthermore, we present two different approaches to represent the actions taken by the policy. The first approach considers continuous actions as soft labels, while the other uses discrete actions as hard assignment of labels for unlabeled examples. We validate the effectiveness of the proposed method on two public benchmark datasets as well as one e-commerce dataset. The results show that the proposed method is able to consistently outperform state-of-the-art methods in various settings. PU learning refers to the problem of learning from a dataset where only a subset of examples are positively labeled and the rest are not annotated at all. It is a critical task due to its prevalence in various real-world applications [1], [2], [3]. In many common situations only positive data are available, for instance, an e-commerce website may only record users who have clicked on advertisements or purchased items. Meanwhile, it is not possible to simply assume that unlabeled instances are negative.

Predicting delayed outcomes is an important problem in recommender systems (e.g., if customers will finish reading an ebook). We formalize the problem as an adversarial, delayed online learning problem and consider how a proxy for the delayed outcome (e.g., if customers read a third of the book in 24 hours) can help minimize regret, even though the proxy is not available when making a prediction. Motivated by our regret analysis, we propose two neural network architectures: Factored Forecaster (FF) which is ideal if the proxy is informative of the outcome in hindsight, and Residual Factored Forecaster (RFF) that is robust to a non-informative proxy. Experiments on two real-world datasets for predicting human behavior show that RFF outperforms both FF and a direct forecaster that does not make use of the proxy. Our results suggest that exploiting proxies by factorization is a promising way to mitigate the impact of long delays in human-behavior prediction tasks.

Woodie Flowers SM '68, MEng '71, PhD '73, the Pappalardo Professor Emeritus of Mechanical Engineering, passed away on Oct. 11 at the age of 75. Flowers' passion for design and his infectious kindness have impacted countless engineering students across the world. Flowers was instrumental in shaping MIT's hands-on approach to engineering design education, first developing teaching methods and learning opportunities that culminated in a design competition for class 2.70, now called 2.007 (Design and Manufacturing I). This annual MIT event, which has now been held for nearly five decades, has impacted generations of students and has been emulated at universities around the world. Flowers expanded this concept to high school and elementary school students, working to help found the world-wide FIRST Robotics Competition, which has introduced millions of children to science and engineering.

Artificial intelligence (AI) is rapidly transforming and improving the ways that industries like healthcare, banking, energy, and retail operate. However, there is one industry in particular that offers incredible potential for the application of AI technologies: education. The opportunities -- and challenges -- that the introduction of artificial intelligence could bring to higher education are significant. Today's colleges and universities face a wide range of challenges, including disengaged students, high dropout rates, and the ineffectiveness of a traditional "one-size-fits-all" approach to education. But when big data analytics and artificial intelligence are used correctly, personalized learning experiences can be created, which may in turn help to resolve some of these challenges.