If you indulge in the occasional television crime drama like most of us do, you've no doubt seen some high-tech investigative tools in action. Incredibly sophisticated video facial recognition is one that's commonly featured: Investigators watch live, grainy security footage, zoom in on a suspect's face, instantly snap it into high-resolution and immediately match the face to a criminal's photo in a massive database. But is that super-advanced level of technology realistic? While real-time video facial recognition remains in its infancy, the development of deep learning techniques -- part of the machine learning family -- is advancing the technology at a rapid pace. Deep learning has also fueled advances in a host of other artificial intelligence and cognitive computing applications for intelligence agencies.

The various manifestations of machine learning, deep learning, neural networks, cognitive computing, image recognition, speech recognition and natural language processing are consistently aiding the enterprise in analytic endeavors associated with big data. In many instances, AI is an immediate solution for the volumes and velocities for which big data is known. Touted as one of the primary expressions of big data in the subsequent decade, its emergence should become much more apparent in the coming 12 months largely due to the maturing influence of AI and the cloud itself. Its most eminent application could very well be provisioning a prototype for security measures to truly fortify the IoT. The year 2017 will see additional organizations experimenting with ways those capabilities render big data less daunting and perhaps even more enjoyable.

Apple has stayed true to its promise and published its first academic paper on artificial intelligence. The world's most valuable company has traditionally kept its AI research private but earlier this month Ruslan Salakhutdinov, director of AI research at Apple, made a pledge to start being more open. The new Apple paper -- published December 22 and titled "Learning from simulated and unsupervised images through adversarial training" -- gives an insight into some of the techniques that Apple is using to develop AI. In the study, which was published through the Cornell University Library, Apple researchers explain a technique that can be used to improve how an algorithm learns to "see" what is in an image. The paper's six authors state that using synthetic images (such as those seen in a video game), as opposed to real-world images, can be more efficient when it comes to training AI models known as neural networks, which are designed to think in the same way as the human brain. Because synthetic image data is already labelled and annotated while real-world images aren't.

Downtown Havana resident Margarita Marquez says she received a special Christmas gift this year: web access at home, a rarity in Cuba, a country with one of the lowest internet penetration rates in the world. Ms Marquez, a 67-year-old retired university professor, was among those selected by the government two weeks ago to participate in a pilot project bringing the web into the homes of 2,000 inhabitants of the historic centre of the island's capital. Most of Communist-ruled Cuba's 11.2 million inhabitants only have access to internet at wi-fi hotspots, and only then if they can afford the 80p hourly tariff that represents around five per cent of the average monthly state salary. Only five per cent of Cubans are estimated to enjoy internet at home, which requires government permission. This is usually granted mainly to academics, doctors and intellectuals.

What are lift, KPI, robustness, model fitting, design of experiments, and the 80/20 rule? Answer: KPI stands for Key Performance Indicator, or metric, sometimes called feature. A robust model is one that is not sensitive to changes in the data. Design of experiments or experimental design is the initial process used (before data is collected) to split your data, sample and set up a data set for statistical analysis, for instance in A/B testing frameworks or clinical trials. The 80/20 rules means that 80 percent of your income (or results) comes from 20 percent of your clients (or efforts). What are collaborative filtering, n-grams, Map Reduce, and cosine distance?

To top this, many practical applications of Machine Learning are enabled by Deep Learning that extends the overall field of Artificial Intelligence. Deep Learning breaks down tasks in ways that makes all kinds of machine assists seem likely. At present, deep learning has moved beyond academic applications and is finding its way into our daily lives. Everything we discussed - Driverless cars, better preventive healthcare, even better movie recommendations, are all here today and will only improve given the rapid rate of advancement. Deep learning avoids the necessity of human-coded features and instead incorporates the feature engineering, feature selection, and model fitting into one step. Feature engineering & selection are fundamental to any application of Deep Learning that you can think off.

When it comes to hardware design language, non-Apple/-Samsung phones tend to be all over the place, even within the same product line. One model of a phone may use onscreen soft buttons, the next one may have hard physical buttons. One year's phone may be made of metal with fingerprint sensor on the back, the next it might be leather with a front scanner. Meizu, on the other hand, has stuck with the same design language over at least a half dozen phones released in the past two to three years. The result has been mostly good -- Meizu's phones are very nicely built metal unibody devices with a sturdy oval all-in-one home button that doubles as a fast fingerprint sensor.

LONDON – Drones delivering blood and medicine to rural areas of Tanzania could help to save the lives of many mothers and newborn babies in a country where one of the biggest causes of maternal deaths is blood loss during childbirth, the U.K. aid department said. The Department for International Development (DFID), which has given funding for the trial due to start early next year, said the drone deliveries could assist more than 50,000 births a year in the East African country. The drones will be able to carry up to 1 kg (2 pounds) of medical supplies and reduce delivery times to 19 minutes from the 110 minutes it takes on average by vehicle. "The U.K. is at the forefront of investing in cutting-edge technology to tackle the global challenges of today such as disease pandemics, medical emergencies and disaster responses," said Priti Patel, U.K.'s international development secretary. "This innovative, modern approach ensures we are achieving the best results for the world's poorest people and delivering value for money for British taxpayers," she said in a statement Thursday.

Thank you all for the huge response to this emerging course! We are delighted to have over 2300 students in over 102 different countries and for the overwhelmingly positive and thoughtful reviews. It's such a privilege to share this important topic with everyday people in a clear and understandable way. In this introductory course, the "Backyard Data Scientist" will guide you through wilderness of Machine Learning for Data Science. Accessible to everyone, this introductory course not only explains Machine Learning, but where it fits in the "techno sphere around us", why it's important now, and how it will dramatically change our world today and for days to come. We'll then explore the past and the future while touching on the importance, impacts and examples of Machine Learning for Data Science: To make sense of the Machine part of Machine Learning, we'll explore the Machine Learning process: Our final section of the course will prepare you to begin your future journey into Machine Learning for Data Science after the course is complete.

So far, we have described the application of neural networks to supervised learning, in which we have labeled training examples. An autoencoder neural network is an unsupervised learning algorithm that applies backpropagation, setting the target values to be equal to the inputs. The autoencoder tries to learn a function \textstyle h_{W,b}(x) \approx x. In other words, it is trying to learn an approximation to the identity function, so as to output \textstyle \hat{x} that is similar to \textstyle x. The identity function seems a particularly trivial function to be trying to learn; but by placing constraints on the network, such as by limiting the number of hidden units, we can discover interesting structure about the data.