In the fall of 2010, I traveled to New Zealand, and one of the places I visited was the small south island city of Christchurch. I was charmed by the tree-lined Avon River, the English-style cathedral in the main square, and the mountains looming in the distance. Inside the cathedral was a stack of poems with a moving message of peace. I saved one to tack on my cork board at home, where it remains to this day. Three months later I turned on the news to see the Christchurch cathedral splintered and broken, its spire crumbled to the ground.

Mandarin Chinese is a notoriously difficult language to translate to English, and for those who rely on Google Translate to decipher important information, machine-based tools simply aren't good enough. All that is about to change, as Google today announced it has implemented a new learning system in its web and mobile translation apps that will bring significantly better results. As a native speaker (and reader and writer) of both Mandarin Chinese (both complex and traditional alphabets) and English, I've often cringed at Google Translate's output. But looking at the examples provided by Google on its blog post, I am impressed. The new system uses what the company calls Google Neural Machine Translation (GNMT), which looks at entire sentences as it decodes instead of breaking them up into words and phrases to be considered independently. The latter method often resulted in disjointed results that sometimes didn't make sense.

The formidable processing power and analytical tools available in public clouds could make industrial IoT more effective and less expensive. But bringing IoT data into the cloud takes more than a network connection. On Tuesday, two companies moved to help enterprises adapt their IoT data for popular cloud services. OSIsoft introduced its PI Integrator for Microsoft Azure, and Particle announced a custom integration with Google Cloud Platform. While some large enterprises with sensitive IoT data do all their analytics in-house, public clouds offer greater scale and better security than many organizations can achieve on their own, MachNation analyst Dima Tokar said.

This post is by Herain Oberoi, Senior Director of Product Marketing at Microsoft. Just over a year ago, we announced the Microsoft Cortana Intelligence Suite with the goal of helping customers use Big Data, Advanced Analytics and Machine Learning technologies in the cloud to derive insights, predict outcomes and prescribe actions based on their data. Cortana Intelligence provides an end to end set of cloud services and pre-configured solutions to enable line of business departments such as Sales, Marketing, Finance, Operations and more to accelerate their digital transformation projects. A year into the release, Cortana Intelligence has had big impact on customers across healthcare, retail, manufacturing, financial services and government and other industries. The range of creative solutions that our customers have come up with, to address their business needs, has been very impressive.

Model ensembling is a very powerful technique to increase accuracy on a variety of ML tasks. In this article I will share my ensembling approaches for Kaggle Competitions. For the first part we look at creating ensembles from submission files. The second part will look at creating ensembles through stacked generalization/blending. I answer why ensembling reduces the generalization error. Finally I show different methods of ensembling, together with their results and code to try it out for yourself. This is how you win ML competitions: you take other peoples' work and ensemble them together." The most basic and convenient way to ensemble is to ensemble Kaggle submission CSV files. You only need the predictions on the test set for these methods -- no need to retrain a model. This makes it a quick way to ensemble already existing model predictions, ideal when teaming up. Let's see why model ensembling reduces error rate and why it works better to ensemble low-correlated model ...

In a world filled with data, being a digital business is no longer enough. To meet market expectations, leading companies are infusing digital intelligence into the equation. This is what we call cognitive business. Unleash your company's cognitive potential at IBM World of Watson 2016, where we'll equip you with the capacity to extract knowledge from data, enhance personal expertise, and outthink the needs of the market at amazing speeds. Become a cognitive business, and see how data, analytics, and Watson can change your world.

Google debuted its smart messaging app Allo last week, resulting in the usual rush of online reviews, denunciations, and proclamations. But whether you think this particular app is successful or not, one thing is clear: this next generation of smart communication apps will change our industry in a big way. The notion of a social feed is only a decade old, but combined with basic messaging apps (text, iMessage, Whatsapp), the modern member of society has a near constant stream of data to manage. The marriage of artificial intelligence to these platforms promises to deliver us a new paradigm of communicating with the world around us. This new form of AI won't be a substitute for our own brain power, but rather a form of social adjacent intelligence to help us respond and post faster, funnier, smarter, more relevant, pithier, and more effectively. In essence, to live a better digital life.

CommAI-env (Environment for Communication-based AI) is a platform for training and evaluating AI systems as described in A Roadmap towards Machine Intelligence. CommAI-env is a platform for training and testing an AI system, the Learner (coded in an arbitrary language of the system developer's choice), in a communication-based setup where it interacts via a bit-level interface with an Environment. The Environment asks the Learner to solve a number of communication-based Tasks, and assigns it a Reward for each task instance it successfully completes. The Learner is presented, in random order, with multiple instances of all tasks, and it has to solve as many of them as possible in order to maximize reward. Examples of tasks currently implemented include counting problems, tasks where the Learnes must memorize lists of items and answer questions about them, or follow navigation instructions through a text-based navigation scheme (see this document for detailed descriptions of the tasks). The set of tasks is open: we are constantly extending it, and we invite others to contribute.

Microsoft CEO Satya Nadella: "We're infusing intelligence into everything we deliver." Microsoft has announced several new artificial intelligence-powered features for Office 365 and Dynamics 365 for Sales. At its annual Ignite conference this week in Atlanta, GA, Microsoft is showing off new AI features across its core products for the enterprise. According to CEO Satya Nadella, this AI-everywhere strategy amounts to Microsoft "democratizing AI" for everyone to solve the world's most pressing challenges. Machine learning, task automation and robotics are already widely used in business.

Google is one of the leading providers of artificial intelligence-assisted language translation, and the company now says a new technique for doing so is vastly improving the results. The company's AI team calls it the Google Neural Machine Translation system, or GNMT, and it initially provided a less resource-intensive way to ingest a sentence in one language and produce that same sentence in another language. Instead of digesting each word or phrase as a standalone unit, as prior methods do, GNMT takes in the entire sentence as a whole. "The advantage of this approach is that it requires fewer engineering design choices than previous Phrase-Based translation systems," writes Quoc V. Le and Mike Schuster, researchers on the Google Brain team. When the technique was first employed, it was able to match the accuracy of those existing translation systems.