Researchers at the Massachusetts Institute of Technology (MIT) have published the first database of annotated English sentences written by non-native English speakers to help train natural language processing systems. English is the most-used language on the Internet, but the majority of people who use English online are non-native speakers, which can make it difficult for language processing algorithms to analyze large amounts of text with the imperfect grammatical quirks non-native speakers often exhibit. The database consists of 5,124 English sentences written by native speakers of 10 different languages, all of which contain at least one grammatical error, along with annotation about the parts of speech used and the relationship between different words, and a corrected version of each sentence for comparison.

Mark Palmer is senior vice president and general manager of engineering at TIBCO. Much has been made about the business implications of recent, rapid advancements in cognitive computing -- that is, the possibility of advanced analytics tools to help human knowledge workers glean actionable insight from vast and deep lakes of historical, transactional and machine-generated information. When utilized well, cognitive tools help humans identify patterns and surface previously undetected cyberattack patterns on your company, customer buying behavior or predictive signals of catastrophic equipment failure based on readings from sensor-enabled devices. But as your business inevitably becomes more algorithmic, you're faced with the next problem: Many algorithms, once discovered, have a remarkably short shelf-life. Algorithmic excellence in analytics requires more than just great math; you must also become as agile at killing off weak or vanquished algorithms as a NASCAR pit crew changing worn tires -- you need to be replacing them with promising new ones.

Intel (NASDAQ:INTC) told technology developers few days ago that the company is planning to bring to the market a new version of its Xeon Phi processor (code named Knights Mill) next year with special features supporting artificial intelligence ("AI"). The ultimate goal of scientists with AI is making the computer to mimic human brain. Since the basic advantage of a human brain over a computer is that the human brain can perform extensive parallel processing, primary requirement for running AI workloads is massive amount of parallel processing. The question is can the Xeon Phi of future offer this kind of parallel processing? According to a recent research conducted by researchers at the University of Massachusetts Dartmouth, modern budget GPUs are capable of offering similar application boosts compared to high-end GPUs in HPC (high performance computing) despite having lower floating point precision capabilities.

The growth of the digital economy has brought about different waves of technology, with the next one being a wave of connectivity. Computing has moved from the traditional desktop models to series of networks known as the Internet of Things (IOT). What this shift in paradigm means for us is that many of the objects that surrounds us will be on a network in one form or another. This recent development has doubled the research on sensor network technologies and advancement in Radio Frequency Identification (RFID), to enable the invisible connection of information and communications systems. Driven by the prevalence and proliferation of smart devices enabled by open wireless technologies such as Wi-Fi, Bluetooth, and RFID, the IOT is an evolution of the internet as we know it, into a network of interconnected equipment, objects and systems.

There's one recurring obsession that keeps haunting almost every marketing and ad agency executive. It's in nearly every Powerpoint deck and on every marketer's mind… Millennials: the oh so documented and ubiquitous M Word!. For many years, most marketers (and their agencies) relied on the same old recipe. Brand awareness and reach were the only metrics that mattered. Problem is, the recipe started to turn sour.

You are sat in a café thinking about a last minute birthday present for a friend or relative who you will be meeting later that day. Using your smartphone, you find something suitable via Amazon or similar web-based retailer. You place the order, paying a slight premium for instant delivery. Twenty minutes later you receive a notification that a drone will be arriving shortly at a delivery-port close to the café. You make the five minutes' walk to the designated meeting point.

Nvidia CEO Jen-Hsun Huang said the partnership illustrates the commitment both companies have made to advancing the use-cases of AI. Nvidia and Chinese search engine giant Baidu are teaming up to develop a cloud-based platform for use in artificially intelligent, self-driving cars. The partnership combines Nvidia's self-driving computing platform with Baidu's cloud and mapping technology to develop an algorithm-based operating system capable of powering complex navigation systems in autonomous vehicles. The open platform will be available for branded car OEM consumer vehicle offerings, as well as fleets of driverless commercial vehicles. The companies are going after Level 3 autonomous vehicle control, which means the car can drive itself (and park itself) in some circumstances, but with a competent human driver at the ready.

Artificial intelligence could improve the interpretation of lung function tests for the diagnosis of long-term lung diseases, according to the findings of a new study. The results, presented today (04 September, 2016) at the European Respiratory Society's International Congress, are the first to explore the potential use of artificial intelligence for improving the accuracy of the diagnosis of lung diseases. Current testing requires a series of methods including a spirometry test, which measures the amount (volume) and the speed (flow) of air during breathing, followed by a body plethysmography test measuring static lung volumes and airways resistance and finally a diffusion test, which measures the amount of oxygen and other gases that cross the lungs' air sacs. Analysis of the results of these tests is largely based on expert opinion and international guidelines, attempting to detect a pattern in the findings. In this new study, researchers included data from 968 people who were undergoing complete lung function testing for the first time.

In an earlier post, I described work that I had initially done as an Insight Data Engineering Fellow. That work, now merged into Flink's master branch, was to do an efficient exact k-nearest neighbors (KNN) query using quadtrees. I have since worked on an approximate version of the KNN algorithm, and I will discuss one method I used for the approximate version using Z-value based hashing. For large and high dimensional data sets, an exact k-nearest neighbors query can become infeasible. There are many algorithms that reduce the dimensionality of the points by hashing them to lower dimensions.

This is the second part of Tamis van der Laan's article featured in the new DZone Guide to Big Data Processing, Volume III. Get your free copy for more insightful articles, industry statistics, and more. So far we have assumed we only have a machine learning model, a training set of samples, and a optimization algorithm to learn from these examples. The next thing we will talk about is the problem of overfitting. If we take our example of a discriminative classifier, we see that it splits the space into two distinct regions for each class.