In this post on integrating SigOpt with machine learning frameworks, we will show you how to use SigOpt and TensorFlow to efficiently search for an optimal configuration of a convolutional neural network (CNN). There are a large number of tunable parameters associated with defining and training deep neural networks ( Bergstra [1]) and SigOpt accelerates searching through these settings to find optimal configurations. This search is typically a slow and expensive process, especially when using standard techniques like grid or random search, as evaluating each configuration can take multiple hours. SigOpt finds good combinations far more efficiently than these standard methods by employing an ensemble of state-of-the-art Bayesian optimization techniques, allowing users to arrive at the best models faster and cheaper. In this example, we consider the same optical character recognition task of the SVHN dataset as discussed in a previous post.

Brain science draws legions of eager students to the field and countless millions in dollars, euros and renminbi to fund research. These endeavors, however, have not yielded major improvements in treating patients who suffer from psychiatric disorders for decades. The languid pace of translating research into therapies stems from the inherent difficulties in understanding mental illness. "Psychiatry deals with brains interacting with the world and with other brains, so we're not just considering a brain's function but its function in complex situations," says Quentin Huys of the Swiss Federal Institute of Technology (E.T.H. Zurich) and the University of Zurich, lead author of a review of the emerging field of computational psychiatry, published this month in Nature Neuroscience. Computational psychiatry sets forth the ambitious goal of using sophisticated numerical tools to understand and treat mental illness.

A novel written through an artificial intelligence program has passed through the first round of judging for the Hoshi Shinichi Literary Award. The book is called, The Day A Computer Writes A Novel, and is one of 11 books submitted to the prize that were written by robots (1,450 books were submitted in total). "I was surprised at the work because it was a well-structured novel. But there are still some problems [to overcome] to win the prize, such as character descriptions," said Satoshi Hase, a Japanese science fiction novelist who was part of the press conference surrounding the award. These programs are not without some human intelligence.

There is growing concern across the board that we might be losing control over cybersecurity. The rapid changes in how we use technology to communicate and the increased number of connected devices means the points of entry or breach are growing. Because the pace of change has been so rapid, security hasn't adapted fast enough and hackers are taking full advantage. The traditional ways of dealing with cyber threats are beginning to look hopelessly inadequate. This concern goes right to the top.

General Motors Co. GM -1.12 % 's proposed purchase of tiny Cruise Automation Inc. for more than 1 billion would be one of the auto industry's biggest Silicon Valley acquisitions to date. And it would certainly not be the last. Auto makers and car-parts suppliers have hooked into tech startups in recent years to boost their in-vehicle connectivity and to accelerate autonomous-car development. The Bay Area already is dotted with auto-industry outposts funding or recruiting Silicon Valley engineering talent to keep pace with Alphabet Inc. GOOGL -0.36 % 's Google X and others working on self-driving cars. Carol Reiley, president of Bay Area autonomous-driving startup Drive.ai, said her company, which like Cruise Automation and Zoox has been operating under the radar, just raised 12 million from venture-capital investors.

The folks over at x.ai – creators of Amy, the artificial intelligence answer to scheduling meetings – have had a shot at showing exactly what it looks like inside their bot's brain, using AI, of course. The team used a powerful deep-learning model, a Recurrent Neural Network (RNN), to trawl 500,000 words in its database, looking at their sequence in a sentence to understand what they mean, then predicting how to categorize them. This year's edition of TNW Conference in Amsterdam includes some of the biggest names in tech. Without a human ever telling the RNN the definitions of different word groups, it has managed to understand that Stanford is different from Instagram, and that Jesse, Luke and Jason are names. This data was cut to down to the 3,500 most frequently used words and has then been projected into a 2D shape in order to show the relationships the AI has made between different words.

The Associated Press-NORC Center for Public Affairs Research poll on drugs and substance abuse was conducted by NORC Feb. 11-14. It is based on online and telephone interviews of 1,042 adults who are members of NORC's nationally representative AmeriSpeak panel. The original sample was drawn from respondents selected randomly from NORC's National Frame based on address-based sampling and recruited by mail, email, telephone and face-to-face interviews. NORC interviews participants over the phone if they don't have Internet access. With a probability basis and coverage of people who can't access the Internet, surveys using AmeriSpeak are nationally representative.

For decades robots have won the affection of humans in pop culture -- whether it was R2-D2 in Star Wars or Rosie the Robot in the Jetsons. But as popular as these machines have been on screen, we haven't seen similar robots enter our everyday lives. Now that appears on the verge of changing as the technology sector increasingly invests in robotics. Funding to private robotics companies doubled in 2015 to reach a record high of 587 million, according to CB Insights. And robots were the dominant theme at a conference Amazon held earlier this week to encourage inspiration and creativity in tech's hottest fields.

The District of Columbia is infamous for some of the political issues that make it distinct from other areas of the country. Its residents only have a non-voting delegate in Congress, for example. And its crippled subway system is uniquely hobbled by the fact that it relies on money from Maryland and Virginia, not just funding from riders and D.C.'s government. So it's no surprise to learn that not long from now, D.C. residents may be able to add drone delivery to their "left out on" list. Many online shoppers are waiting eagerly for the day that they'll be able to order something on Amazon.com and have it dropped off, via drone, on their front stoop.

The Federal Aviation Administration is predicting a bright future for the growth of the commercial and hobbyist drone industries after final regulations are approved. In an aerospace forecast report released Thursday, the FAA said unmanned aircraft systems will be the "most dynamic growth sector within aviation." It noted that venture capitalists have already sunk "considerable" funds into the industry in hopes of building early market share. Already, the FAA predicted that 1.9 million hobbyist drones will be sold this year, along with more than 600,000 commercial drones. The FAA predicts that 4.3 million hobbyist drones could be sold per year by 2020.