Tests for Esophageal cancer can be expensive, uncomfortable and can have side effects. For many patients, we can predict non-existence of disease with 100% certainty, just using demographics, lifestyle, and medical history information. Our objective is to devise a general methodology for customizing tests using user preferences so that expensive or uncomfortable tests can be avoided. We propose to use classifiers trained from electronic health records (EHR) for selection of tests. The key idea is to design classifiers with 100% false normal rates, possibly at the cost higher false abnormals. We compare Naive Bayes classification (NB), Random Forests (RF), Support Vector Machines (SVM) and Logistic Regression (LR), and find kernel Logistic regression to be most suitable for the task. We propose an algorithm for finding the best probability threshold for kernel LR, based on test set accuracy. Using the proposed algorithm, we describe schemes for selecting tests, which appear as features in the automatic classification algorithm, using preferences on costs and discomfort of the users. We test our methodology with EHRs collected for more than 3000 patients, as a part of project carried out by a reputed hospital in Mumbai, India. Kernel SVM and kernel LR with a polynomial kernel of degree 3, yields an accuracy of 99.8% and sensitivity 100%, without the MP features, i.e. using only clinical tests. We demonstrate our test selection algorithm using two case studies, one using cost of clinical tests, and other using "discomfort" values for clinical tests. We compute the test sets corresponding to the lowest false abnormals for each criterion described above, using exhaustive enumeration of 15 clinical tests. The sets turn out to different, substantiating our claim that one can customize test sets based on user preferences.

This paper investigates the effectiveness of factorial speech processing models in noise-robust automatic speech recognition tasks. For this purpose, the paper proposes an idealistic approach for modeling state-conditional observation distribution of factorial models based on weighted stereo samples. This approach is an extension to previous single pass retraining for ideal model compensation which is extended here to support multiple audio sources. Non-stationary noises can be considered as one of these audio sources with multiple states. Experiments of this paper over the set A of the Aurora 2 dataset show that recognition performance can be improved by this consideration. The improvement is significant in low signal to noise energy conditions, up to 4% absolute word recognition accuracy. In addition to the power of the proposed method in accurate representation of state-conditional observation distribution, it has an important advantage over previous methods by providing the opportunity to independently select feature spaces for both source and corrupted features. This opens a new window for seeking better feature spaces appropriate for noisy speech, independent from clean speech features.

As recently as 2013, the [deep learning] space saw fewer than 10 deals. Computer Vision: Startups here are using deep learning for image recognition, analytics, and classification. Aerial image analytics startup Terraloupe was seed-funded this year by Germany-based Bayern Kapital. New York-based Calrifai -- backed by investors including Google Ventures, Lux Capital, and NVidia -- entered the R/GA accelerator this year, after raising 10M in Series A in Q2'15. Captricity, which extracts information from hand-written data, has raised 49M in equity funding so far from investors including Social Capital, Accomplice, White Mountains Insurance Group, and New York Life Insurance Company.

Humanoid robots could soon allow us to stream our consciousness anywhere in the world, acting as surrogate bodies that enable people to'instantly be in multiple places at once.' This is the goal of the Avatar XPrize, an ANA-sponsored concept for a 22 million contest just selected at the XPrize Visioneers 2016 Summit as'ready to launch.' The technology would make James Cameron's Avatar a reality, developing'limitless travel' avatars that can be rented and controlled remotely by a human operator, who will be able to hear, see, and feel what the robot is experiencing. Humanoid robots could soon allow us to stream our consciousness anywhere in the world, acting as surrogate bodies that enable people to'instantly be in multiple places at once.' The Avatar XPrize aims to facilitate the creation of'avatars that you – the public will be able to use to travel anywhere, anytime, instantly.'

Originally published in the ITS Ghaziabad 2nd CXO Meet Souvenir, "Digital India Mission: Transforming India for Tomorrow." Achievements in machine learning are coming at ever-increasing rapidity over the past several months. You are likely familiar with the recent accomplishments associated with machine learning, especially those of so-called deep learning, or the use of multi-layered artificial neural networks. These specific achievements include the high profile AlphaGo and Deep Dream, along with numerous others in the realms of computer vision and natural language processing. Interestingly, a number of these recent mainstream successes are primarily attributable to Google in one form or another.

With dozens of public satellites in orbit and many more scheduled over the next decade, the size and complexity of geospatial imagery continues to grow. It has become increasingly difficult to manage this flood of data and use it to gain valuable insights. That's why we're excited to announce that we're bringing two of the most important collections of public, cost-free satellite imagery to Google Cloud: Landsat and Sentinel-2. The Landsat mission, developed under a joint program of the USGS and NASA, is the longest continuous space-based record of Earth's land in existence, dating back to 1972 with the Landsat 1 satellite. Landsat imagery sets the standard for Earth observation data due to the length of the mission and the rich data provided by its multispectral sensors.

Artificial intelligence driven by deep learning often runs on many computer chips working together in parallel. But the deep-learning algorithms, called neural networks, can run only so fast in this parallel computing setup because of the limited speed with which data flows between the different chips. The Japan-based multinational Fujitsu has come up with a novel solution that sidesteps this limitation by enabling larger neural networks to exist on a single chip. The neural networks used in deep learning typically run on graphics processing units (GPUs) that originated as components for generating and displaying images. By creating an efficiency shortcut in the calculations performed by neural networks, Fujitsu researchers reduced the amount of internal GPU memory used by 40 percent.

With an aim to lead the world of smartphones with its artificial intelligence (AI)-based technology, Google on Tuesday launched much-awaited Pixel -- a new premium device completely designed by the tech giant with Google Assistant built right-in -- at a special event here. Don't Miss: Black Friday 2016: Everything You Need to Know Now The launch also ended the Nexus branding under which the company has always released phones in partnership with other original equipment manufacturers like LG (for Google Nexus 5) and Huawei (for Google Nexus 6P). Although HTC has manufactured the smartphones, the new device bears Google branding. With curved sculpted edges and a unibody made up of combination of aerospace grade aluminum and glass, the device comes in two sizes -- 5 and 5.5-inch with 2.5D Corning Gorilla Glass 4 protected super AMOLED display. Pixel is available in two -- quite black and very silver colors in India.

Artificial intelligence has captured the rhythm of science fiction. For example, the script of a new science fiction short is the creation of a bot. Although the software provides the order of the word choices, the source material is human. It works by algorithm and it derives its poetic power from the words of human feeling. The results are surprisingly good and even funny, in spite of its mechanized origins.

Japanese auto firm Denso, a subsidiary of Toyota, displays a robotic limb designed to support a surgeon's arm Meanwhile, Sharp is taking aim at the housing market with pint-sized Rin-chan, which can operate home appliances based on its owners' feelings. For example, if a house dweller says'it's too hot', the robot will turn on the air conditioning. Another star of the show is a mug-sized, doe-eyed robot called Kirobo Mini made by Toyota as a chatty companion for its human owners.