Machine reading comprehension (MRC) is an AI challenge that requires machine to determine the correct answers to questions based on a given passage. MRC systems must not only answer question when necessary but also distinguish when no answer is available according to the given passage and then tactfully abstain from answering. When unanswerable questions are involved in the MRC task, an essential verification module called verifier is especially required in addition to the encoder, though the latest practice on MRC modeling still most benefits from adopting well pre-trained language models as the encoder block by only focusing on the "reading". This paper devotes itself to exploring better verifier design for the MRC task with unanswerable questions. Inspired by how humans solve reading comprehension questions, we proposed a retrospective reader (Retro-Reader) that integrates two stages of reading and verification strategies: 1) sketchy reading that briefly investigates the overall interactions of passage and question, and yield an initial judgment; 2) intensive reading that verifies the answer and gives the final prediction. The proposed reader is evaluated on two benchmark MRC challenge datasets SQuAD2.0 and NewsQA, achieving new state-of-the-art results. Significance tests show that our model is significantly better than the strong ALBERT baseline. A series of analysis is also conducted to interpret the effectiveness of the proposed reader.

This week Bett, the education show that brings together over 800 education providers, takes center stage in London. Educators, developers, and ecosystem players come together to share what is new, connect and learn. Microsoft is the worldwide partner for Bett, but most platform providers and hardware vendors use the event to launch their latest devices and software solutions aimed at education. As in years past, we have announcements aimed at making life in the classroom easier for the teacher, whether it is about saving time on managing students, assets, or content. Microsoft added new indicator lights at the back of the computers the students are using so teachers can quickly glance at the class and make sure all machines are powered and connected.

As mass shootings at US schools increase in frequency while our country's gun control laws remain weaker than those in any other developed nation, more school administrators across the US are turning to artificially intelligent surveillance tools in an attempt to beef up school safety. But systems that allow schools to easily track people on campus have left some worried about the impact on student privacy. Recode has identified at least nine US public school districts -- including the district home to Marjory Stoneman Douglas High School (MSD) in Parkland, Florida, which in 2018 experienced one of the deadliest school shootings in US history -- that have acquired analytic surveillance cameras that come with new, AI-based software, including one tool called Appearance Search. Appearance Search can find people based on their age, gender, clothing, and facial characteristics, and it scans through videos like facial recognition tech -- though the company that makes it, Avigilon, says it doesn't technically count as a full-fledged facial recognition tool. Even so, privacy experts told Recode that, for students, the distinction doesn't necessarily matter.

Signal temporal logic (STL) is an expressive language to specify time-bound real-world robotic tasks and safety specifications. Recently, there has been an interest in learning optimal policies to satisfy STL specifications via reinforcement learning (RL). Learning to satisfy STL specifications often needs a sufficient length of state history to compute reward and the next action. The need for history results in exponential state-space growth for the learning problem. Thus the learning problem becomes computationally intractable for most real-world applications. In this paper, we propose a compact means to capture state history in a new augmented state-space representation. An approximation to the objective (maximizing probability of satisfaction) is proposed and solved for in the new augmented state-space. We show the performance bound of the approximate solution and compare it with the solution of an existing technique via simulations.

Estimates suggest that only 20 per cent of today's engineers are employable in this age of new technologies like artificial intelligence (AI), internet of things (IoT), blockchain and cyber security. And it's high time that educational institutions, businesses and the government upped their game. These are the findings of a report unveiled by the BML Munjal University, a higher education institution promoted by the Hero Group. The report, titled ÁI & Future of Work: Redefining Future of Enterprise, analyses the opportunities and challenges brought about by new-age tech changes and presents a roadmap for academic institutions, enterprises as well as the government on how to work together to fulfill the demand for qualified professionals in this new age where exponential technologies like AI and blockchain are going to rule the roost. "Today, legacy skills, tools and technologies have become obsolete. New-age digital professionals proficient in AI, IoT are being called upon to enter the talent workforce, with a new set of skills," said Sameer Dhanrajani, CEO of AIQRATE Advisory, who authored the report.

Created by Lazy Programmer Inc., Lazy Programmer Team Comment Policy: Please write your comments according to the topic of this page posting. Comments containing a link will not be displayed before approval.

Uber has been one of the most active companies trying to accelerate the implementation of real world machine learning solutions. Just this year, Uber has introduced technologies like Michelangelo, Pyro.ai and Horovod that focus on key building blocks of machine learning solutions in the real world. This week, Uber introduced another piece of its machine learning stack, this time aiming to short the cycle from experimentation to product. PyML, is a library to enable the rapid development of Python applications in a way that is compatible with their production runtime. The problem PyML attempts to address is one of those omnipresent challenges in large scale machine learning applications.

The New York Institute of Finance (NYIF) and Google Cloud announced a new Machine Learning for Trading Specialization available exclusively on the Coursera platform. The Specialization helps learners leverage the latest AI and machine learning techniques for financial trading. Amid the Fourth Industrial Revolution, nearly 80 percent of financial institutions cite machine learning as a core component of business strategy and 75 percent of financial services firms report investing significantly in machine learning. The Machine Learning for Trading Specialization equips professionals with key technical skills increasingly needed in the financial industry today. Composed of three courses in financial trading, machine learning, and artificial intelligence, the Specialization features a blend of theoretical and applied learning.

After a school shooting in Parkland, Florida left 17 people dead, RealNetworks decided to make its facial recognition technology available for free to schools across the US and Canada. If school officials could detect strangers on their campuses, they might be able to stop shooters before they got to a classroom. Anxious to keep children safe from gun violence, thousands of schools reached out with interest in the technology. Dozens started using SAFR, RealNetworks' facial recognition technology. From working with schools, RealNetworks, the streaming media company, says it's learned an important lesson: Facial recognition isn't likely an effective tool for preventing shootings.

Artificial intelligence offers a way to solve that problem.