Deep neural networks that learn to represent data in multiple layers of increasing abstraction have dramatically improved the state-of-the-art for speech recognition, object recognition, object detection, predicting the activity of drug molecules, and many other tasks. Deep learning discovers intricate structure in large datasets by building distributed representations, either via supervised, unsupervised or reinforcement learning.

Big data engineers often face a conceptual task – how to extract knowledge from the information stored. For a long period of time our company has been developing natural language processing (NLP) methods of analyzing and mining data (http://www.intellexer.com/). In our practice, we faced storages with millions of manually created documents/reports and the lack of possibilities for systemizing the information obtained at different times, in different places and without a uniform standard for presenting data. There is no single and easy answer. Unfortunately, it's impossible to offer a universally applicable method that would satisfy all the customer's needs.

Intel, long the most dominant chip vendor for PCs, is making a high-profile transition away from them and toward such growth markets as the internet of things (IoT), drones, the cloud, artificial intelligence (AI) and machine learning, virtual and augmented reality, network connectivity and the data center. Essentially, the company wants the billions of connected devices that will make up the IoT to run on Intel technology, and for its products to drive the systems that connect them to the cloud. "We are transforming into the company that is powering the cloud, connecting smart devices and making new experiences possible based on all that today," Intel CEO Brian Krzanich told the more than 6,000 people who traveled to San Francisco last week for the annual Intel Developer Forum (IDF). In the past, the show had been packed with new PCs and servers powered by the company's latest and greater processors. However, this year's IDF had a decidedly different vibe to it, with fewer PCs on display and more drones, cars and other connected devices on the floor.

AI has been making a lot of progress lately by almost any standard. It has quietly become part of our world, powering markets, websites, factories, business processes and soon our houses, our cars and everything around us. But the biggest recent successes have also come with surprising failures. Tesla impressed the world by launching a self driving car, but then crashed in cases a human would have easily handled. AlphaGo beat the human champion Go player years before most experts thought possible, but completely collapsed after its opponent played an unusual move.

This is the year of the chatbot. With the rise of artificial intelligence (AI), bots are responding to queries, helping users find products, and even troubleshooting customer service issues. Artificial intelligence dealmaking has exploded recently, leaping to a new quarterly record of over 140 deals in Q1â 16. This year especially, artificial intelligence (AI) has had a renaissance â " Tesla pushed their self-driving autopilot out to all eligible cars, and Google and Facebook have both announced large investments in AI research. The race is on to get driverless trucks on the roads, and experts say the impact on professional drivers â is going to be hugeâ .. read more.

And you probably notice countless other details as well. A machine sees none of this; an image is encoded as an array of numbers representing color values for each pixel, as in the second photo, the one on the right. So how do we enable machine vision to go from pixels to a deeper understanding of an image?

But a machine sees none of this; to a machine, it's just a bunch of pixels. It's up to computer-vision technology like the one developed at FAIR to segment each object out. Considering that real-world objects come in so many shapes and sizes as well as the fact that photos are subject to varying backgrounds and lighting conditions, it's easy to see why visual recognition is so complex. The answer, Dollar writes, lies in deep convolutional neural networks that are "trained rather than designed." The networks essentially learn from millions of annotated examples over time to identify the objects. "The first stage would be to look at different parts of the image that could be interesting," he says. "The second step is to then say, 'OK, that's a sheep,' or'that's a dog.' "Our whole goal is to get at all the pixels, to get at all the information in the image," he says.

Assigning people the tedious task of manually reviewing and manipulating documents is a sure-fire way to miss errors and introduce inaccuracies as blurry-eyed employees move between programs and cut-and-paste data. Although this is standard practice at many companies, it's a problem that Robotic Process Automation (RPA) technology is now able to solve. RPA can mimic a human's ability to gather data and put data into Excel spreadsheets. It can access disparate systems (ERP, sales, tax, etc.), review tens of thousands of documents in an hour or two (instead of several days), do so without error, and put the data in a standardized format. But RPA's advantages go far beyond imitating keystrokes and avoiding errors; RPA's speed, accuracy and formatting abilities make it an essential foundational technology for a more automated, intelligent enterprise.

The world's largest social network is getting better at learning about your photographs, but it still needs humans to help it improve. Facebook announced Thursday that it will make its computer vision research available to anyone who wants it, ideally creating a brainstorm that advances the technology that can automatically identify objects in a photograph. "We're making the code for DeepMask SharpMask as well as MultiPathNet -- along with our research papers and demos related to them -- open and accessible to all, with the hope that they'll help rapidly advance the field of machine vision," Facebook research scientist Piotr Dollar wrote in a blog post. If you're scratching your head, think of it this way: You probably deal with this technology on a regular basis if you're a Facebook user, even if you don't realize it. When you upload a picture and try to tag someone, Facebook will automatically suggest that person's name unless they've disabled the "tag suggestions" feature.

What is machine learning, when and how is it used? Why is it such a buzzword? Listen to Josefin Rosén explaining and giving examples how machine learning make sense of all data we have and how to create value of it.