Recognizing activities of daily living (ADLs) in the real world is an important task for understanding everyday human life. However, even though our life events consist of chronological ADLs with the corresponding places and objects (e.g., drinking coffee in the living room after making coffee in the kitchen and walking to the living room), most existing works focus on predicting individual activity labels from sensor data. In this paper, we introduce a novel framework that produces an event timeline of ADLs in a home environment. The proposed method combines semantic concepts such as action, object, and place detected by sensors for generating stereotypical event sequences with the following three real-world properties. First, we use temporal interactions among concepts to remove objects and places unrelated to each action. Second, we use commonsense knowledge mined from a language resource to find a possible combination of concepts in the real world. Third, we use temporal variations of events to filter repetitive events, since our daily life changes over time. We use cross-place validation to evaluate our proposed method on a daily-activities dataset with manually labeled event descriptions. The empirical evaluation demonstrates that our method using real-world properties improves the performance of generating an event timeline over diverse environments.

This paper provides a review and commentary on the past, present, and future of numerical optimization algorithms in the context of machine learning applications. Through case studies on text classification and the training of deep neural networks, we discuss how optimization problems arise in machine learning and what makes them challenging. A major theme of our study is that large-scale machine learning represents a distinctive setting in which the stochastic gradient (SG) method has traditionally played a central role while conventional gradient-based nonlinear optimization techniques typically falter. Based on this viewpoint, we present a comprehensive theory of a straightforward, yet versatile SG algorithm, discuss its practical behavior, and highlight opportunities for designing algorithms with improved performance. This leads to a discussion about the next generation of optimization methods for large-scale machine learning, including an investigation of two main streams of research on techniques that diminish noise in the stochastic directions and methods that make use of second-order derivative approximations.

The digital workplace will fundamentally change the way we work. Established and emerging technologies play a big part in this, but our day to day work and the way organizations manage this work is undergoing a radical shift. Gartner VP Matthew W Cain and Gartner research director Helen Poitevin outlined 11 emerging trends that will shape digital workplaces in the years to come at the Gartner Digital Workplace Summit in London earlier this year, many of which held true. But as Cain and Poitevin noted, the digital workplace is not just about technologies. The following 10 trends explore the how the way we work will continue to change over the coming year.

Movie buffs have been hearing about artificial intelligence for years – from Steven Spielberg's 2001 science fiction drama AI to the 2015 robotic police force in Chappie and beyond. AI is no longer the stuff of science fiction. This essential part of the technology sector aims to create intelligent machines of all kinds that think, work and react like humans. Just as electricity transformed the way industries functioned in the past century, artificial intelligence -- the science of programming cognitive abilities into machines -- has the power to substantially change society in the next 100 years. AI is being harnessed to enable such things as home robots, robo-taxis and mental health chatbots to make you feel better.

Machine learning (aka A.I.) seems bizarre and complicated. It's the tech behind image and speech recognition, recommendation systems, and all kinds of tasks that computers used to be really bad at but are now really good at. It involves teaching a computer to teach itself. And you can learn to do it in well under a year, according to data scientist Bargava. You'll need to put in a solid 10-20 hours a week, but you will learn a lot along the way.

Thanks to advances in AI and machine learning, a slow but steady transformation is coming to education -- under the hood. When you compare the typical 21st century classroom with that of the early 1900s, the differences aren't terribly obvious. Teachers will be standing in front, giving instructions and sharing notes on a modern-day version of the old blackboard -- say, an overhead projector or a shared computer display. Students will be sitting at their desks in the classroom or watching via online video-conferencing software. The technology has changed: A lot of the tools and processes have been digitized, some of it has been automated, and geographical barriers have been removed to some extent -- but the actors and elements have remained much the same.

This post is authored by Chris Testa-O'Neill, Applied Data Scientist in the Microsoft Cloud AI team. Artificial Intelligence (AI) is proving to be a massively disruptive force, one that is leading to the digital transformation of businesses at a faster pace than most of us would have imagined. At Microsoft, our mission is to bring AI to every developer and every organization on the planet, and to provide the best platform and tools to make them successful. You can read more Microsoft's approach to AI here. In keeping with our mission, we are currently running a series of popular AI boot camps around the world.

A woman glumly considers her chances of expanding knowledge and skills for a rapidly changing world at the facilities of TCDC Resource Center, one spot to get help. The Fourth Industrial Revolution is at our doorstep, with artificial intelligence (AI), robotics and automation threatening unskilled workers and new graduates who are at risk of being left unemployed by job-killing androids. As these technological breakthroughs take place, automation and AI are compelling blue-collar workers to improve their skills and do more sophisticated work. Those who fail to adapt will be left behind. Plans to cut jobs have already begun in the banking and telecom sectors, signalling a trend that retail may soon follow as customers enjoy self-service through digital channels.

While it's always possible to compute a variational approximation to a posterior distribution, it can be difficult to discover problems with this approximation". We propose two diagnostic algorithms to alleviate this problem. The Pareto-smoothed importance sampling (PSIS) diagnostic gives a goodness of fit measurement for joint distributions, while simultaneously improving the error in the estimate.

Big numbers are nothing new to Quizlet, one of the most widely-used digital study tools in the United States. And according to one website tracker, it is among the most visited sites in the country (besting LinkedIn and Spotify). But there's another number that also has the company excited. This week Quizlet raised an additional $20 million in a Series B round led by Icon Ventures. Other investors include Union Square Ventures, Costanoa Ventures, Owl Ventures and Altos Ventures.