15th Dec.' 17 - Headlines

Google Brain Co-founder Andrew Ng wants to expand artificial intelligence beyond traditional IT world

Andrew Ng starts “Landing.ai” to bring AI into manufacturing, Foxconn joins in

Andrew Ng has announced a new startup “Landing.ai” that focuses on using artificial intelligence into the manufacturing sector. Under the venture, Foxconn has become the .first strategic partner. Noting that the fruits of AI and machine learning should reach every industry, instead of being confined to internet companies, Ng said his initiative Landing.ai is working on a number of “AI transformation programs” that include introducing new technologies to companies and training employees,among others. As for the Foxconn partnership, Ng said that he’s been working with them to develop “AI technologies, talent and systems that build on the core competencies of the two companies.”

BlockSci v0.3 released

BlockSci version 0.3 announced with big performance gains, SegWit support, and numerous bugfixes

BlockSci, a high-performance platform for blockchain science and exploration, has released its version 0.3 that includes many bug fixes and a massive 5x performance improvement, in addition to SegWit support. Arvind Naraynan, who co-created BlockSci at Princeton University, said that in the updated version traversing every transaction input and output on the bitcoin blockchain takes only 1 second. “If you want to start using BlockSci immediately, we have made available an EC2 image: ami-7cf38706. We recomend using instance with 60 GB of memory or more for optimal performance (r4.2xlarge). On boot, a Jupyter Notebook running BlockSci will launch immediately,” the platform founders said, “To run BlockSci locally, you must be running a full node (such as bitcoind or an altcoin node) since BlockSci requires the full serialized blockchain data structure which full nodes produce.” Details are available in the release notes.

Hadoop 3 released

Announcing Apache Hadoop v3.0.0

The Apache Software Foundation has announced the general availability of Apache Hadoop v3.0.0 with new features that improve the efficiency, scalability, and reliability of the open source framework. "This latest release unlocks several years of development from the Apache community," said Chris Douglas, Vice President of Apache Hadoop. "The platform continues to evolve with hardware trends and to accommodate new workloads beyond batch analytics, particularly real-time queries and long-running services. At the same time, our Open Source contributors have adopted Apache Hadoop to a wide range of deployment environments, including the Cloud."

Among the highlights, the new version has improved capabilities for cloud storage systems such as Amazon S3 (S3Guard), Microsoft Azure Data Lake, and Aliyun Object Storage System. The HDFS erasure coding halves the storage cost of HDFS while also improving data durability, whereas the YARN Timeline Service v.2 (preview) improves the scalability, reliability, and usability of the Timeline Service. "Hadoop 3 is a major milestone for the project, and our biggest release ever," said Andrew Wang, Apache Hadoop 3 release manager. The team announced that Apache Hadoop will be in action at the Strata Data Conference in San Jose, CA, on 5-8 March 2018.

Building a distributed deep learning environment in hours

Announcing free trials for IBM Spectrum Conductor Deep Learning Impact version 1.1

Built on IBM Spectrum Conductor with Spark, the IBM Spectrum Conductor Deep Learning Impact provides robust, end-to-end workflow support for deep learning application logic. This includes the complete lifecycle management from installation and configuration, data ingest and preparation, building, optimizing, and training the model, to inference and testing.

The IBM Spectrum Conductor Deep Learning Impact V1.1 adds on to IBM Spectrum Conductor with Spark V2.2.1 that provides deep learning capabilities to a new or existing IBM Spectrum Conductor with Spark environment. It is supported with IBM Power servers with NVLink and NVIDIA GPUs. It is also supported on x86 systems leveraging GPUs.

The IBM Spectrum Conductor Deep Learning Impact helps monitor deep learning training jobs by capturing logs from the underlying deep learning frameworks. It also provides training progress of a job with monitoring data charts and summary information, and what more, offers advices on how to optimize the training job. The full version of IBM Spectrum Conductor Deep Learning Impact 1.1 is available on IBM Passport Advantage. To evaluate and try IBM Spectrum Conduct Deep Learning Impact, download the free trial.

Advancing the Industrial Internet of Things (IIoT)

Industrial Internet Consortium and Plattform Industrie 4.0 collaborate to facilitate interoperability of Industrial Internet systems

Industrial Internet Consortium and Plattform Industrie 4.0 have aligned their internal architectures to accelerate the adoption of the Industrial Internet on a global scale with a cross-industry oriented approach. Representatives from both organizations began meeting in 2015, independently developing their own reference architectures for the Industrial Internet: Plattform Industrie 4.0’s Reference Architecture Model for Industrie 4.0 (RAMI4.0) and the Industrial Internet Reference Architecture (IIRA). After subsequent discussions, the companies merged the two approaches creating an initial draft mapping showing the direct relationships between elements of the models. They jointly developed a clear roadmap to ensure future interoperability. More information will be available in coming days.

Microsoft rolls out new Artificial Intelligence features to Office 365

Microsoft introduces “Insights” in Excel, “Acronyms” on Word, “Time to leave” on Outlook, Microsoft Whiteboard Preview, and “Text in image search” features into Office 365 suite

Microsoft has announced the preview of Insights in Excel—a new service that automatically highlights patterns it detects, which makes it easier for everyone to explore and analyze their data. Powered by machine learning, Insights helps identify trends, outliers, and other useful visualizations, providing new and useful perspectives on data.

Next, the company announced a new Word feature called Acronyms. Using machine learning, Acronyms helps people understand shorthand that is commonly used in their own workplaces by leveraging the Microsoft Graph to surface definitions of terms that have been previously defined across emails and documents. Acronyms will begin rolling out to Word Online for Office 365 commercial subscribers in 2018.

In another new addition, Microsoft is bringing Cortana to the Outlook mobile app such that when it is time to leave for appointments, Outlook will now send a notification—with directions for both driving and public transit—taking into account current location, the event location, and real-time traffic information. Time to leave in Outlook is rolling out to iOS users this month in markets where Cortana is available.

Microsoft also announced the preview of Microsoft Whiteboard for Windows 10 devices—a freeform digital canvas where people, ideas, and content can come together. Microsoft Whiteboard Preview is built for teams who ideate and work together across multiple devices and locations.It is  now available in preview from the Windows Store.

Finally, Microsoft has launched a new Text in images feature with intelligent search. The new feature will “automatically extract searchable text” from whiteboards, screenshots, receipts, business cards and more. Text in image search is currently rolling out and will be available to all Office 365 commercial subscribers by the end of December.

“AI Factory” in France

Microsoft’s "AI Factory" initiative may boost the emergence of artificial intelligence market in France

Microsoft has created the "AI Factory" program at Station F, the largest start-up campus in the world hosting around 1,000 startups in Paris. Designed to  "stimulate the emergence of French champions of artificial intelligence," the program supports seven start-ups: Recast.AI, AB Tasty, DCbrain, Scortex, AI Craft, Case Law Analytics, Hugging Face. Microsoft is specially focusing on leveraging artificial intelligence in the marketing sector, using Chatbots, CRM, predictive analysis, and more. The company said that it has launched the program in order to help all players in the French Tech space to address the shift in artificial intelligence to make France a world reference in the field.

In other Data science News

ARM releases pretrained TensorFlow models and code for their speech keyword recognition code

ARM has released pretrained TensorFlow models and code for their speech keyword recognition code which is small enough to run on M-class DSPs. The details are published in this arXiv paper. The code, model definitions and pretrained models are available here on GitHub.

China develops three year action plan to accelerate the industrial development of artificial intelligence

China's Ministry of Industry and Information Technology has released the three year action plan (2018-2020) to promote the development of a new generation of artificial intelligence. The move is aimed to boost the "Made in China 2025 " and "a new generation of artificial intelligence development plan" which China wants to integrate into its mainstream economy in coming years.