Training is the make-or-break task in every development project that involves artificial intelligence (AI). Determining an AI application’s fitness for its intended use involves training it with data from the solution domain into which it will be deployed.
In 2018, developers will come to regard training as a potential bottleneck in the AI application-development process and will turn to their AI solution providers for robust training tools. Developers will adopt robust tools for training AI models for disparate applications and deployment scenarios. By the end of this coming year, AI model training will emerge as the fastest growing platform segment in big data analytics. To keep pace with growing developer demand, most leading analytics solution providers will launch increasingly feature-rich training tools.
During the year, we’ll see AI solution providers continue to build robust support for a variety of AI-model training capabilities and patterned pipelines in their data science, application development, and big-data infrastructure tooling. Many of these enhancements will be to build out the automated ML capabilities in their DevOps tooling. By year-end 2018, most data science toolkits will include tools for automated feature engineering, hyperparameter tuning, model deployment, and other pipeline tasks. At the same time, vendors will continue to enhance their unsupervised learning algorithms to speed up cluster analysis and feature extraction on unlabeled data. And they will expand their support for semi-supervised learning in order to use small amounts of labeled data to accelerate pattern identification in large, unlabeled data sets.
In 2018, synthetic (aka artificial) training data, will become the lifeblood of most AI projects. Solution providers will roll out sophisticated tools for creation of synthetic training data and the labels and annotations needed to use it for supervised learning.
The surge in robotics projects and autonomous edge analytics will spur solution providers to add strong reinforcement learning to their AI training suites in 2018. This will involve building AI modules than can learn autonomously with little or no “ground truth” training data, though possible with human guidance. By the end of the year, more than 25 percent of enterprise AI app-dev projects will involve autonomous edge deployment, and more than 50 percent of those projects will involve reinforcement learning.
During the year, more AI solution providers will add collaborative learning to their neural-net training tools. This involves distributed AI modules collectively exploring, exchanging, and exploiting optimal hyperparameters so that all modules may converge dynamically on the optimal trade-off of learning speed vs. accuracy. Collaborative learning approaches, such as population-based training, will be a key technique for optimizing AI in that’s embedded in IoT&P (Internet of Things and People) edge devices.
It will also be useful in for optimizing distributed AI architectures such as generative adversarial networks (GANs) in the IoT, clouds, or even within server clusters in enterprise data centers. Many such training scenarios will leverage evolutionary algorithms, in which AI model fitness is assessed emergently by collective decisions of distributed, self-interested entities operating from local knowledge with limited sharing beyond their neighbor entities.
Another advanced AI-training feature we’ll see in AI suites in 2018 is transfer learning. This involves reuses of some or all of the training data, feature representations, neural-node layering, weights, training method, loss function, learning rate, and other properties of a prior model. Typically, a developer relies on transfer learning to tap into statistical knowledge that was gained on prior projects through supervised, semi-supervised, unsupervised, or reinforcement learning. Wikibon has seen industry progress in using transfer learning to reuse the hard-won knowledge gained in training one GAN on GANs in adjacent solution domains.
Also during the year, edge analytics will continue to spread throughout into enterprise AI architectures. During the year, edge-node on-device AI training will become a standard feature of mobile and IoT&P development tools. Already, we see it in many leading IoT and cloud providers’ AI tooling and middleware.
About the author: About the author: James Kobielus is a Data Science Evangelist for IBM. James spearheads IBM’s thought leadership activities in data science. He has spoken at such leading industry events as IBM Insight, Strata Hadoop World, and Hadoop Summit. He has published several business technology books and is a very popular provider of original commentary on blogs and many social media.
The post Developers Will Adopt Sophisticated AI Model Training Tools in 2018 appeared first on Datanami.
Read more here:: www.datanami.com/feed/
By Erin Scherer
2017 was an amazing year here at TeamARIN! We collaborated with guest bloggers from the community and subject matter experts inside ARIN to bring you lots of great content. As we look forward to a brand new year, we wanted to take a moment to look back at some of the most popular posts from 2017.
We’re always looking to improve what we do here and we wanted to share what we’re doing right – according to you! So, we rounded up the top ten posts our readers found to be the most helpful or interesting from this past year on TeamARIN. Enjoy!
16 Years with IPv6 – Torbjörn Eklöv, Co-Founder and Owner of Interlan, was an early adopter of IPv6. He shares the benefits of IPv6 adoption and offers advice to anyone who wants to enable IPv6 in their network with a static configuration from an ISP.
Tech industry Takes Strides Toward IPv6 – Looking back over the past year, we’ve seen some exciting news of increased IPv6 activity from major tech companies. While 2015 marked the year the ARIN region ran out of IPv4, 2016 marked the year more companies made a push toward its successor.
12 Steps to Enable IPv6 in an ISP Network – Jordi Palet Martinez summarizes the 12 fundamental steps necessary to achieve native IPv6 support and maintain IPv4 as a transparent service.
IPv6 is Not Optional – Mattias Lindgren, Senior Network Engineer at the University of Colorado Denver, details how he spearheaded the effort to enable IPv6 across two campuses and explains why adopting IPv6 is no longer optional.
Stop Procrastinating and Do IPv6 – Mike Milne of Carleton University takes us through the IPv6 planning process, highlighting why working on a complete end-to-end network design is enjoyable.
Let’s Talk About Quick and Easy Way to SWIP – James Ricewick, Resource Analyst, talks about a new, quick and easy method of submitting reassignments from your ARIN Online account.
Stay on the Cutting Edge with IPv6 – Rob Carsey explains how Monmouth University went from IPv6 zero to IPv6 hero in less than one summer and why IPv6 is important to stay on the cutting edge in the education field.
Using the Market to Obtain or Release IPv4 – A few pointers on how to navigate the IPv4 transfer market, whether you are interested in obtaining or releasing IPv4 address space.
Why Does DNS Security Matter? – Securing DNS is critical to ensuring online safety. We discuss the importance of DNSSEC and share the services ARIN offers to help you secure your reverse zones.
Implementing RPKI: It’s Easier Than You Think – Andrew Gallo, Principal IT Architect and Network Engineer at George Washington University, explains the importance of implementing RPKI and why it may be easier than you think.
We hope that if you don’t already have IPv6, you are including it in your plans for 2018. We have plenty of resources to help you achieve your organization’s goal in the new year. Check out our Get6 page to learn more. And if you are interested in contributing a guest post on TeamARIN, get in touch with us. We love collaborating with our readers and community. Happy New Year!
Read more here:: teamarin.net/feed/
By Andrew Brust
Click to learn more about video blogger Andrew Brust. The Big Data & Brews video blog series continues with host Andrew Brust, Senior Director of Market Strategy and Intelligence at Datameer. The series touches on hot topics within the business of Big Data, Analytics, Internet of Things, Machine Learning, Cloud Computing, Modern BI, NoSQL and Next […]
The post How Can Companies Balance Between Too Much Control and Too Much Access to Data? appeared first on DATAVERSITY.
Read more here:: www.dataversity.net/feed/
Click to learn more about author Sreeram Sreenivasan. From autonomous cars to messaging apps to IoT devices, every piece of technology has come to be powered by data. No matter what your profession or job title, you’re bound to encounter some form of data on a daily basis. However, the volume of data is not as […]
Read more here:: www.dataversity.net/feed/
By Ken Hosac
Click to learn more about author Ken Hosac. The Internet of Things (IoT) is much more than just connecting devices to the Internet and Cloud, it’s about generating new business insights, automating business and production processes and accelerating innovation cycles. The vast array of IoT implementations are difficult to comprehend, as they can encompass everything from […]
Read more here:: www.dataversity.net/feed/