We describe a novel family of models of multi- layer feedforward neural networks in which the activation functions are encoded via penalties in the training problem. Our approach is based on representing a non-decreasing activation function as the argmin of an appropriate convex optimiza- tion problem. The new framework allows for algo- rithms such as block-coordinate descent methods to be applied, in which each step is composed of a simple (no hidden layer) supervised learning problem that is parallelizable across data points and/or layers. Experiments indicate that the pro- posed models provide excellent initial guesses for weights for standard neural networks. In addi- tion, the model provides avenues for interesting extensions, such as robustness against noisy in- puts and optimizing over parameters in activation functions.

Modern decision analytics frequently involves the optimization of an objective over a finite horizon where the functional form of the objective is unknown. The decision analyst observes covariates and tries to learn and optimize the objective by experimenting with the decision variables. We present a nonparametric learning and optimization policy with covariates. The policy is based on adaptively splitting the covariate space into smaller bins (hyper-rectangles) and learning the optimal decision in each bin. We show that the algorithm achieves a regret of order $O(\log(T)^2 T^{(2+d)/(4+d)})$, where $T$ is the length of the horizon and $d$ is the dimension of the covariates, and show that no policy can achieve a regret less than $O(T^{(2+d)/(4+d)})$ and thus demonstrate the near optimality of the proposed policy. The role of $d$ in the regret is not seen in parametric learning problems: It highlights the complex interaction between the nonparametric formulation and the covariate dimension. It also suggests the decision analyst should incorporate contextual information selectively.

The framework of reinforcement learning or optimal control provides a mathematical formalization of intelligent decision making that is powerful and broadly applicable. While the general form of the reinforcement learning problem enables effective reasoning about uncertainty, the connection between reinforcement learning and inference in probabilistic models is not immediately obvious. However, such a connection has considerable value when it comes to algorithm design: formalizing a problem as probabilistic inference in principle allows us to bring to bear a wide array of approximate inference tools, extend the model in flexible and powerful ways, and reason about compositionality and partial observability. In this article, we will discuss how a generalization of the reinforcement learning or optimal control problem, which is sometimes termed maximum entropy reinforcement learning, is equivalent to exact probabilistic inference in the case of deterministic dynamics, and variational inference in the case of stochastic dynamics. We will present a detailed derivation of this framework, overview prior work that has drawn on this and related ideas to propose new reinforcement learning and control algorithms, and describe perspectives on future research.

The researchers found that machine learning -- the science of getting computers to learn over time without human intervention -- is as accurate as a team of child and adolescent psychiatrists, including a forensic psychiatrist, in determining risk for school violence. "Previous violent behavior, impulsivity, school problems and negative attitudes were correlated with risk to others," says Drew Barzman, MD, a child forensic psychiatrist at Cincinnati Children's Hospital Medical Center and lead author of the study. "Our risk assessments were focused on predicting any type of physical aggression at school. We did not gather outcome data to assess whether machine learning could actually help prevent school violence. That is our next goal."

This article first appeared in Data Sheet, Fortune's daily newsletter on the top tech news. To get it delivered daily to your in-box, sign up here. Andrew Ng, the respected computer scientist and entrepreneur, is getting his hands dirty, metaphorically speaking. Having co-founded education startup Coursera and built artificial intelligence units for Google (googl) and then Baidu, Ng now has his sights on factories. "I've been to so many manufacturing plants," says Ng, whose new startup, Landing.ai,

You'll even discover how to use artificial neural networks and deep learning structures for classification!

In A.I., data is power. A.I. algorithms and how they are plugged into each other is an art that is becoming known through university courses, online training, and literally by people watching YouTube videos. Artificial Intelligence is open source, and it should be. What you can do to protect your company from competition is build proprietary datasets. There are plenty of datasets open to the public.

The role that adult education and colleges play in preparing the labour market for technological disruption will be explored by MPs. The World Economic Forum, which holds an annual conference at the Swiss ski resort of Davos, set the theme of its 2016 gathering of world leaders around the topic of the fourth industrial revolution. Since then, the term has entered common parlance and it is now widely characterised as concerning what impact that new technologies, including artificial intelligence and robotics, will have on the labour market with many low and medium skilled jobs believed to be at risk of automation. Now the education select committee is to explore the issue of preparing for the so-called fourth industrial revolution. Stressing the importance of the inquiry, committee chairman Robert Halfon said by the 2030s, as many as 28 per cent of the current jobs taken by 16- 24-year-olds are likely to be at risk of automation.

The pressure is on for companies and governments to address the ways that artificial intelligence (AI) is altering the future of work. In this video, recorded at the Aspen Ideas Festival in June, experts--Markle Foundation CEO and president Zoë Baird; Joy Buolamwini, founder of the Algorithmic Justice League at MIT Media Lab; James Fallows, national correspondent of the Atlantic; and Coursera cofounder Andrew Ng--discuss how to make the transition into this new age easier for everyone. Andrew Ng: AI is the new electricity. About 100 years ago, we started rolling out electricity in the United States, and it changed every single major industry, everything ranging from healthcare and culture to transportation, communications, and manufacturing are now all electricity powered. We now see a surprisingly clear path for AI to also transform every single major industry.

AI, or Artificial Intelligence, is cropping up more and more in eLearning conversations--who's using it and how, and what it means for the future of corporate digital learning. As Learning Solutions prepares to explore AI from many angles, an overview of foundational aspects of AI might come in handy. Here, we'll introduce concepts and trends that are likely to appear in any deep discussion of using AI in eLearning. Artificial intelligence refers broadly to technologies that can learn and perform specific tasks. More complex tasks entail machine learning, a next-level technology that takes an AI machine or technology and teaches it to make "decisions" based on algorithms, learn from those decisions, and refine its own performance.