Real-Time Machine Learning (RTML): DARPA Throws Down the Gauntlet

by topic_admin

The DARPA Grand Challenge, subsequently reborn as the Urban Challenge, the Robotics Challenge and the FANG Challenge, has always been about pushing the boundaries of technologies. Whether it’s about autonomous vehicles racing through the desert or robots working in hazardous environments, the core objective is to achieve the impossible – yet, at the least, the highly improbable.

This year, the DARPA Grand Challenge brings machine learning to the forefront, however on a real-time basis. The Real-time Machine Learning application, which places DARPA and the National Science Foundation in partnership, is roughly “the creation of a processor that can proactively interpret and learn from data in real-time, solve unfamiliar problems using what it has learned, and operate with the energy efficiency of the human brain.” Funding is defined at $10 million to the winner.

What is Real-Time Machine Learning?

Image result for real-time machine learning

Real-time machine learning is not a brand new field, however it&# 1 8217;s largely limited to language recognition and computer vision. Some of the other applications include credit card acceptance in checkout time, which generally must occur within a couple of seconds, recommendation engines with real-time inputs, routing in logistics and transportation, etc.

True real-time learning involves removing the bottlenecks which are traditionally existing in now ’s processing, including ETL (extract-transform-load) batch procedures that invariably render the data obsolete even before the analysis can occur.

One solution for this issue of processing bottlenecks is the utilization of in-memory computing integrated with multilayer perceptron profound learning attributes. This can typically by scaled for operational data collections in the petabytes. One such system is the GridGain Continuous Learning Framework, which is intended for companies that don’t have massive budgets for machine learning jobs. It is built on Apache Ignite, which includes an in-memory database, in-memory data grid and streaming analytics. It accelerates NoSQL and relational databases, and provides high availability and horizontal scalability while being consistent with distributed SQL.

This particular DARPA Grand Challenge is tasked with finding these real-time machine learning methods, but people which are essentially built from the floor up and may provide low-power and lightweight alternatives for the identical kind of real-time machine learning capabilities.

The barrier is split into two 18-month periods. Phase I expects to eliminate the frequently prohibitive prices of application-specific ICs by creating hardware compilers which will enable ML-specific chips to be automatically generated from high-level source code, even while Phase II requires two presentation applications to be supported by the hardware optimization manufactured in Phase I.

Related Articles

Leave a Comment

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept