Deep learning has provided the world of data science with highly effective tools that can address problems in virtually any domain, and using nearly any kind of data. However, the non-intuitive features deduced and used by deep learning algorithms require a very careful experimental design, and a failure to meet that requirement can lead to miserably flawed results, regardless of the quality of the data or the structure of the deep learning network. I first noticed such flaws almost ten years ago, when I applied algorithms that used non-intuitive features for the purpose of automatic face recognition. I noticed that when using the most common face recognition benchmarks at that time (FERET, ORL, YaleB, JAFFE, and others), the algorithms could identify the correct face even when using just a small seemingly blank part of the background, normally a small sub-image from the top-left corner of the original image, that does not contain any part of the face, hair, clothes, or anything else that could allow the recognition of a person (1). I ran the experiments like they were intended, but instead of using the full face images I used a very small part of the background taken from the top-left corner of each image. The algorithms were able to identify the faces in very high accuracy, sometimes as high as 100%, even though no faces were in the images that were analyzed.

A form of machine learning called deep learning is one of the key technologies behind recent advances in applications like real-time speech recognition and automated image and video labeling. The approach, which uses multi-layered artificial neural networks to automate data analysis, also has shown significant promise for health care: It could be used, for example, to automatically identify abnormalities in patients' X-rays, CT scans and other medical images and data. In two new papers, UCLA researchers report that they have developed new uses for deep learning: reconstructing a hologram to form a microscopic image of an object and improving optical microscopy. Their new holographic imaging technique produces better images than current methods that use multiple holograms, and it's easier to implement because it requires fewer measurements and performs computations faster. The research was led by Aydogan Ozcan, an associate director of the UCLA California NanoSystems Institute and the Chancellor's Professor of Electrical and Computer Engineering at the UCLA Henry Samueli School of Engineering and Applied Science; and by postdoctoral scholar Yair Rivenson and graduate student Yibo Zhang, both of UCLA's electrical and computer engineering department.

Google's DeepMind subsidiary has created a computer system, AlphaZero, that uses deep neural networks to train itself at superhuman speed. It recently taught itself to play chess, Shogi and Go at beyond-world-champion levels in less than one day. The implications of artificial intelligences that are far smarter than we are worrisome, and given how fast AI is developing and how fast its creations can learn new skills, urgent....

Pratik works as a Data Scientist at Synechron. Involved in various innovative projects and concepts, he applies a range of Machine Learning and Deep Learning algorithms to create and deliver business value. He previously was an admired Consulting Staff Principal at Oracle. His creative write-ups have awarded him many prizes over the years from Television 18 Broadcast Ltd. Pratik holds an MBA in Finance with Information Technology from IBS Bangalore.

Previous works have used deep learning. They used regression to predict the colour of each pixel. This, however, produces fairly bland and dull results. Previous works used Mean Squared Error (MSE) as the loss function to train the model. The authors noted that MSE will try to'average' out the colors in order to get the least average error, which will result in a bland look.

In standard supervised machine learning, we need training data, i.e. a set of data points with known labels, and we build a model to learn the distinguishing properties that separate data points with different labels. This trained model can then be used to make label predictions for new data points. If we want to make predictions for another task (with different labels) in a different domain, we cannot use the model trained previously. We need to gather training data with the new task, and train a separate model. Transfer learning provides a framework to leverage the already existing model (based on some training data) in a related domain.

All of a sudden, everyone is talking about artificial intelligence. It appears as frequently in the news headlines as indefinite articles before nouns. Not to speak about all the other over-hyped buzzwords from the tech scene. This guide clears up misconceptions about A.I., deep learning, and machine learning, among others to offer a clear overview for healthcare practitioners. It also describes in what fields A.I. can change medicine and healthcare, as well as present the companies with the best track-records.

Looking for a way to turn your home computer into a deep-learning AI super-monster? Nvidia has an expensive answer. The new Titan V GPU promises a crazy amount of processing for deep learning and AI applications. It's nine times more powerful -- at 110 teraflops -- than last year's Titan X, Nvidia's last massive desktop graphics processor aimed at machine learning applications. The Titan V is based on Nvidia's newer Volta chip architecture, which is also being used in Nvidia's Xavier self-driving car system and for data centers. This is the first desktop-based use of the new chip, and it costs $2,999 or £2,700 (about AU$4,800).

Four hours is all it took for Google's DeepMind artificial intelligence program to learn everything there was to know about chess, The Telegraph reported Wednesday. DeepMind's AlphaZero program, which teaches itself from scratch, achieved "superhuman" knowledge of chess in less than the amount of time you'd spend, say, watching the extended version of The Lord of the Rings: Return of the King. Chess has long been used to test the ability of artificial intelligence because the game's rigid structure is ideal for programming a computer with rules, and then letting it run its own tests against those rules. AlphaZero started this experiment knowing only the basics of chess gameplay, but by playing thousands of games against itself, AlphaZero updated its neural network with information about the effectiveness of certain moves -- over and over again, until it became the best chess player in the known universe. "The games AlphaZero played ... are far beyond anything humans or chess computers have come up with," said David Kramaley, a chess education expert.