This paper considers policy search in continuous state-action reinforcement learning problems. Typically, one computes search directions using a classic expression for the policy gradient called the Policy Gradient Theorem, which decomposes the gradient of the value function into two factors: the score function and the Q-function. This paper presents four results:(i) an alternative policy gradient theorem using weak (measure-valued) derivatives instead of score-function is established; (ii) the stochastic gradient estimates thus derived are shown to be unbiased and to yield algorithms that converge almost surely to stationary points of the non-convex value function of the reinforcement learning problem; (iii) the sample complexity of the algorithm is derived and is shown to be $O(1/\sqrt(k))$; (iv) finally, the expected variance of the gradient estimates obtained using weak derivatives is shown to be lower than those obtained using the popular score-function approach. Experiments on OpenAI gym pendulum environment show superior performance of the proposed algorithm.

As powerful as reinforcement learning is, Dr. LeCun says he believes that other forms of machine learning are more critical to general intelligence. "My money is on self-supervised learning," he said, referring to computer systems that ingest huge amounts of unlabeled data and make sense of it all without supervision or reward. He is working on models that learn by observation, accumulating enough background knowledge that some sort of common sense can emerge. "Imagine that you give the machine a piece of input, a video clip, for example, and ask it to predict what happens next," Dr. LeCun said in his office at New York University, decorated with stills from the movie "2001: A Space Odyssey." "For the machine to train itself to do this, it has to develop some representation of the data. It has to understand that there are objects that are animate and others that are inanimate. The inanimate objects have predictable trajectories, the other ones don't."

'Curse of dimensionality' is a well-known problem in Data Science, which often causes poor performance, inaccurate results, and, most importantly, a similarity measure break-down. The primary cause of this is because high dimensional datasets are typically sparse, and often a lower-dimensional structure or'Manifold' would embed this data. So there is a non-linear relationship among the variables (or features or dimensions), which we need to learn to compute better similarity. Manifold learning is an approach to non-linear dimensionality reduction. The basis for algorithms in manifold learning is that the dimensionality of many data sets is only artificially high 1.

In light of the recent events surrounding Covid-19, learning for grades K-12 looks very different than it did a month ago. Parents and educators may be feeling overwhelmed about turning their homes into classrooms. With that in mind, a team led by Media Lab Associate Professor Cynthia Breazeal has launched aieducation.mit.edu to share a variety of online activities for K-12 students to learn about artificial intelligence, with a focus on how to design and use it responsibly. Learning resources provided on this website can help to address the needs of the millions of children, parents, and educators worldwide who are staying at home due to school closures caused by Covid-19, and are looking for free educational activities that support project-based STEM learning in an exciting and innovative area. The website is a collaboration between the Media Lab, MIT Stephen A. Schwarzman College of Computing, and MIT Open Learning, serving as a hub to highlight diverse work by faculty, staff, and students across the MIT community at the intersection of AI, learning, and education. "MIT is the birthplace of Constructionism under Seymour Papert.

Log-based predictive maintenance of computing centers is a main concern regarding the worldwide computing grid that supports the CERN (European Organization for Nuclear Research) physics experiments. A log, as event-oriented adhoc information, is quite often given as unstructured big data. Log data processing is a time-consuming computational task. The goal is to grab essential information from a continuously changeable grid environment to construct a classification model. Evolving granular classifiers are suited to learn from time-varying log streams and, therefore, perform online classification of the severity of anomalies. We formulated a 4-class online anomaly classification problem, and employed time windows between landmarks and two granular computing methods, namely, Fuzzy-set-Based evolving Modeling (FBeM) and evolving Granular Neural Network (eGNN), to model and monitor logging activity rate. The results of classification are of utmost importance for predictive maintenance because priority can be given to specific time intervals in which the classifier indicates the existence of high or medium severity anomalies.

The recent advances in deep learning indicate significant progress in the field of single image super-resolution. With the advent of these techniques, high-resolution image with high peak signal to noise ratio (PSNR) and excellent perceptual quality can be reconstructed. The major challenges associated with existing deep convolutional neural networks are their computational complexity and time; the increasing depth of the networks, often result in high space complexity. To alleviate these issues, we developed an innovative shallow residual feature representative network (SRFRN) that uses a bicubic interpolated low-resolution image as input and residual representative units (RFR) which include serially stacked residual non-linear convolutions. Furthermore, the reconstruction of the high-resolution image is done by combining the output of the RFR units and the residual output from the bicubic interpolated LR image. Finally, multiple experiments have been performed on the benchmark datasets and the proposed model illustrates superior performance for higher scales. Besides, this model also exhibits faster execution time compared to all the existing approaches.

In multi-label learning, the issue of missing labels brings a major challenge. Many methods attempt to recovery missing labels by exploiting low-rank structure of label matrix. However, these methods just utilize global low-rank label structure, ignore both local low-rank label structures and label discriminant information to some extent, leaving room for further performance improvement. In this paper, we develop a simple yet effective discriminant multi-label learning (DM2L) method for multi-label learning with missing labels. Specifically, we impose the low-rank structures on all the predictions of instances from the same labels (local shrinking of rank), and a maximally separated structure (high-rank structure) on the predictions of instances from different labels (global expanding of rank). In this way, these imposed low-rank structures can help modeling both local and global low-rank label structures, while the imposed high-rank structure can help providing more underlying discriminability. Our subsequent theoretical analysis also supports these intuitions. In addition, we provide a nonlinear extension via using kernel trick to enhance DM2L and establish a concave-convex objective to learn these models. Compared to the other methods, our method involves the fewest assumptions and only one hyper-parameter. Even so, extensive experiments show that our method still outperforms the state-of-the-art methods.

Multilingual sequence labeling is a task of predicting label sequences using a single unified model for multiple languages. Compared with relying on multiple monolingual models, using a multilingual model has the benefit of a smaller model size, easier in online serving, and generalizability to low-resource languages. However, current multilingual models still underperform individual monolingual models significantly due to model capacity limitations. In this paper, we propose to reduce the gap between monolingual models and the unified multilingual model by distilling the structural knowledge of several monolingual models (teachers) to the unified multilingual model (student). We propose two novel KD methods based on structure-level information: (1) approximately minimizes the distance between the student's and the teachers' structure level probability distributions, (2) aggregates the structure-level knowledge to local distributions and minimizes the distance between two local probability distributions. Our experiments on 4 multilingual tasks with 25 datasets show that our approaches outperform several strong baselines and have stronger zero-shot generalizability than both the baseline model and teacher models.

Ordinal peer grading has been proposed as a simple and scalable solution for computing reliable information about student performance in massive open online courses. The idea is to outsource the grading task to the students themselves as follows. After the end of an exam, each student is asked to rank -- in terms of quality -- a bundle of exam papers by fellow students. An aggregation rule then combines the individual rankings into a global one that contains all students. We define a broad class of simple aggregation rules, which we call type-ordering aggregation rules, and present a theoretical framework for assessing their effectiveness. When statistical information about the grading behaviour of students is available (in terms of a noise matrix that characterizes the grading behaviour of the average student from a student population), the framework can be used to compute the optimal rule from this class with respect to a series of performance objectives that compare the ranking returned by the aggregation rule to the underlying ground truth ranking. For example, a natural rule known as Borda is proved to be optimal when students grade correctly. In addition, we present extensive simulations that validate our theory and prove it to be extremely accurate in predicting the performance of aggregation rules even when only rough information about grading behaviour (i.e., an approximation of the noise matrix) is available. Both in the application of our theoretical framework and in our simulations, we exploit data about grading behaviour of students that have been extracted from two field experiments in the University of Patras.

Every day, Massachusetts seventh-grader Kaylyn Wilson takes a break from doing homework online and opens an app on her phone for a half-hour foreign language lesson. "The boy has three green bikes and an egg," the 12-year-old announced to her family in French at the start of her third week using the mobile app from Rosetta Stone, the language-learning software giant. Wilson doesn't yet need to study a language for credit. But during the school shutdowns to contain the coronavirus, her father saw Rosetta Stone advertise free accounts for students – an offer other language-learning software companies have made as well. Wilson decided to give it a go.