Hexagon Capability Center India (HCCI), the research and development unit of Swedish technology major Hexagon AB, will set up an AI community centre in Hyderabad next year in association with Nasscom, a top company official said. The upcoming artificial intelligence (AI) centre will be accessible to all from high-school level, and will provide free education on AI topics, HCCI Vice-President and Country Manager Navaneet Mishra told reporters after presenting awards to winners of a 24-hour hackathon, named Hexathon, which was conducted recently by the company. He said they recently signed an MoU with Nasscom, and the centre will cover about 350 students every year in multiple batches. HCCI, which currently has 1,400-plus employees at its facility in Hyderabad, will touch a 1,500-workforce by the end of this year, Mishra said, to a query on expansion plans. Telangana IT Secretary Jayesh Ranjan said 2020 will be celebrated in Telangana as the'Year of AI' and to have a community-focussed AI initiative supported by Hexagon will be a good activity as a part of the year-long programme.

Discrimination in the U.S. credit market is well documented. Historically, minorities have disproportionately been denied loans, mortgages, and credit cards, or charged higher rates than other customers. Now that artificial intelligence is taking over many credit decisions -- and taking human bias out of the equation -- it'll be easy to enforce laws against discrimination in lending, right? Not necessarily, argues Jann Spiess, an assistant professor of operations, information, and technology at Stanford Graduate School of Business. In a recent paper in The University of Chicago Law Review, he and Talia Gillis, a doctoral student at Harvard Business School and Harvard Law School, examined what happens when existing anti-discrimination rules are applied to choices made by machines.

Stommel, a co-author of An Urgency of Teachers: The Work of Critical Digital Pedagogy (Hybrid Pedagogy, 2018) and a co-founder of the faculty-development event Digital Pedagogy Lab, recently returned to the classroom full time after several years of running Mary Washington's Division of Teaching and Learning Technologies. He spoke with The Chronicle about how professors bring a "full, complicated self" to the classroom, why he thinks students are marginalized, and whether colleges have really gotten serious about teaching. Ten years ago, the student-success conversation was largely about student affairs and financial aid. Now administrators seem to be talking more about the classroom. So are colleges taking teaching more seriously?

Do your skills need an upgrade? Have you heard about machine learning but don't know where to start? Do you want to build your own machine learning projects? Our latest course will help you level up just in time for 2020! Finally, a masterclass that makes machine learning so straightforward that everyone can understand it. The Complete Machine Learning Course for Everybody is your one-stop-shop to go from absolute beginner to machine learning expert.

Learning the different concepts in data science can often feel like a daunting task. Here are 6 books to help lift the burden. This is an almost *exhaustive* book on machine learning topics ranging from the very basics of probability, to mixture models, to variational inference, to deep learning. Even though I first encountered this book as a companion textbook for a university course, I think calling this a textbook is doing it a disservice. It is basically an encyclopedia and can serve as a detailed reference for any data scientist or machine learning engineer.

An easier way for data scientists to build reproducible experiments with machine learning pipelines and communicate operational dependencies to their engineering counterparts as part of a new MLOps approach you deploy to the Cloud and the Edge at scale. Chris Lauren the Principal Program Manager for the Azure Machine Learning Platform goes over the new Azure Machine Learning service. Chris shows you what capabilities data scientists can get across the machine learning lifecycle within a familiar notebook experience. And you'll see how you can use the newly introduced Automated Machine Learning capabilities in Azure ML to build machine learning models in a fraction of the time.

Cross Labs' mission is to bridge between intelligence science and AI technology at the service of human society. At Cross Labs, we focus on pushing fundamental research towards a thorough mathematical understanding of all intelligent processes observable both in nature and in artificial environments. To reach our goals, we are seeking ambitious, highly-skilled researchers to solve open problems on both natural and artificial intelligence fronts. Our current research priorities cover a large range of intelligence science topics, including artificial life, cognitive neuroscience, collective intelligence, deep learning, robotics, and computational linguistics. Other research topics will be seriously considered if you can make a case for their tractability and relevance to intelligence science research as envisioned by Cross Labs.

While the familiar idiom "you say tomayto, I say tomahto" is meant to showcase the triviality of differences, the irony lies in its illustration of the wide variation in English pronunciation. Such vagaries in pronunciation can make English difficult for many nonnative speakers unused to pronouncing certain sounds. English is a stress-based language, meaning that it requires emphasis on particular syllables, said Sarah Daniels, CEO and co-founder of English-learning startup Blue Canoe. "If someone is not proactively thinking about stress … we, in our system, can teach them where it is and how to do it." Bellvue-based Blue Canoe's mobile app directs its users to repeat sentence prompts and record them.

Auto-annotation by ensemble of models is an efficient method of learning on unlabeled data. Wrong or inaccurate annotations generated by the ensemble may lead to performance degradation of the trained model. To deal with this problem we propose filtering the auto-labeled data using a trained model that predicts the quality of the annotation from the degree of consensus between ensemble models. Using semantic segmentation as an example, we show the advantage of the proposed auto-annotation filtering over training on data contaminated with inaccurate labels. Moreover, our experimental results show that in the case of semantic segmentation, the performance of a state-of-the-art model can be achieved by training it with only a fraction (30$\%$) of the original manually labeled data set, and replacing the rest with the auto-annotated, quality filtered labels.

We analyze the practices of reservoir computing in the framework of statistical learning theory. In particular, we derive finite sample upper bounds for the generalization error committed by specific families of reservoir computing systems when processing discrete-time inputs under various hypotheses on their dependence structure. Non-asymptotic bounds are explicitly written down in terms of the multivariate Rademacher complexities of the reservoir systems and the weak dependence structure of the signals that are being handled. This allows, in particular, to determine the minimal number of observations needed in order to guarantee a prescribed estimation accuracy with high probability for a given reservoir family. At the same time, the asymptotic behavior of the devised bounds guarantees the consistency of the empirical risk minimization procedure for various hypothesis classes of reservoir functionals.