DARPA selects research teams to explore paradigm-changing approaches to Machine Learning and Artificial Intelligence

DARPA’s Lifelong Learning Machines (L2M) program, drawing inspiration from biological systems, seeks to develop fundamentally new ML approaches that allow systems to adapt continually to new circumstances without forgetting previous learning.
7 months ago

Machine learning (ML) and artificial intelligence (AI) systems have significantly advanced in recent years. However, they are currently limited to executing only those tasks they are specifically designed to perform and are unable to adapt when encountering situations outside their programming or training. DARPA’s Lifelong Learning Machines (L2M) program, drawing inspiration from biological systems, seeks to develop fundamentally new ML approaches that allow systems to adapt continually to new circumstances without forgetting previous learning.

First announced in 2017, DARPA’s L2M program has selected the research teams who will work under its two technical areas. The first technical area focuses on the development of complete systems and their components, and the second will explore learning mechanisms in biological organisms with the goal of translating them into computational processes. Discoveries in both technical areas are expected to generate new methodologies that will allow AI systems to learn and improve during tasks, apply previous skills and knowledge to new situations, incorporate innate system limits, and enhance safety in automated assignments.

The L2M research teams are now focusing their diverse expertise on understanding how a computational system can adapt to new circumstances in real time and without losing its previous knowledge. One group, the team at University of California, Irvine plans to study the dual memory architecture of the hippocampus and cortex. The team seeks to create an ML system capable of predicting potential outcomes by comparing inputs to existing memories, which should allow the system to become more adaptable while retaining previous learnings. The Tufts University team is examining a regeneration mechanism observed in animals like salamanders to create flexible robots that are capable of altering their structure and function on the fly to adapt to changes in their environment. Adapting methods from biological memory reconsolidation, a team from University of Wyoming will work on developing a computational system that uses context to identify appropriate modular memories that can be reassembled with new sensory input to rapidly form behaviors to suit novel circumstances.

“With the L2M program, we are not looking for incremental improvements in state-of-the-art AI and neural networks, but rather paradigm-changing approaches to machine learning that will enable systems to continuously improve based on experience,” said Dr. Hava Siegelmann, the program manager leading L2M. “Teams selected to take on this novel research are comprised of a cross-section of some of the world’s top researchers in a variety of scientific disciplines, and their approaches are equally diverse.”

While still in its early stages, the L2M program has already seen results from a team led by Dr. Hod Lipson at Columbia University’s Engineering School. Dr. Lipson and his team recently identified and solved challenges associated with building and training a self-reproducing neural network, publishing their findings in Arvix Sanity. While neural networks are trainable to produce almost any kind of pattern, training a network to reproduce its own structure is paradoxically difficult. As the network learns, it changes, and therefore the goal continuously shifts. The continued efforts of the team will focus on developing a system that can adapt and improve by using knowledge of its own structure. “The research team’s work with self-replicating neural networks is just one of many possible approaches that will lead to breakthroughs in lifelong learning,” said Siegelmann.

“We are on the threshold of a major jump in AI technology,” stated Siegelmann. “The L2M program will require significantly more ingenuity and effort than incremental changes to current systems. L2M seeks to enable AI systems to learn from experience and become smarter, safer, and more reliable than existing AI.”

Photo: Today's machine learning and AI systems are limited to executing only tasks they are specifically programmed to perform, without being able to adapt to new situations outside of their training. DARPA's L2M program aims to generate new methodologies that will allow these systems to learn and improve during tasks, apply previous skills and knowledge to new situations, incorporate innate system limits, and enhance safety in automated assignments. (credit: DARPA)

Source: DARPA

Read more

Comments

No comments to display.

Related posts

A Comprehensive IoT Node Proposal Using Open Hardware. A Smart Farming Use Case to Monitor Vineyards

The last decade has witnessed a significant reduction in prices and an increased performance of electronic components, coupled with the influence of the shift towards the generation of open resources, both in terms of knowledge (open access), programs (open-source software), and components (open hardware).

EU's Call for Proposals: Regenerative medicine: from new insights to new applications

Regenerative medicine offers hope for untreatable disease and the ageing population, improved quality of life and reduced medical costs. However, so far, regenerative medicine has not yet proved itself in the clinic beyond rare diseases or conditions of limited public health importance. With recent scientific discoveries opening up new approaches to regenerative medicine, the challenge is to use these to extend the regenerative approach to major diseases and conditions.
Application Deadline in 4 months

Mobile money holds key to financial inclusion in Africa, experts say

More than 10% of adults in sub-Saharan Africa have a mobile money account, the highest in the world, and it is spreading to West Africa and beyond.

EU's Call for Proposals: Innovative approaches to urban and regional development through cultural tourism

The various forms of cultural tourism in Europe are important drivers of growth, jobs and economic development of European regions and urban areas. They also contribute, by driving intercultural understanding and social development in Europe through discovering various types of cultural heritage, to the understanding of other peoples' identities and values. However, although cultural tourism by its nature invites cross border regional and local cooperation, its full innovation potential in this respect is not yet fully explored and exploited.
Application Deadline in 3 months

UMITRON secures US$2 million project funding to deliver IoT and AI technology to drive growth of sustainable aquaculture

IDB Lab's grant resources will be used by UMITRON and Piscis to test and calibrate UMITRON's AI feeding technology. Additionally, funds will be used to both train local producers and scale up UMITRON's solution in the Lake Titicaca region. The end goal is to implement data-driven aquaculture feeding technology for farmers in Peru and eventually all of Latin America.

Call for applications: UN SDG Action Awards

Every year, the UN SDG Action Awards recognize individuals, civil society organizations, subnational governments, foundations, networks, private sector leaders who are advancing the global movement for the Sustainable Development Goals in the most transformative, impactful and innovative way.
Application Deadline in a month

Call for applications: DOEN Foundation Circular Entrepreneurship Programme 2019 for young Social Entrepreneurs

DOEN is looking for sustainable, social entrepreneurs with innovative products or services that will contribute to the creation of this new circular economy. We are looking for people who dare to take risks and seize new opportunities. People who come up with new solutions or business models.
Application Deadline in 2 months

Call for collaborative projects: IoT4Industry - Application of IoT technologies in the manufacturing environment

The IoT4Industry project seeks to support EU growth and competitiveness through the development of a new cross-sectoral industrial value chain based on the integration and use of IoT and related components (sensors, chips, electronics component, communication networks, cloud computing, Big Data, Artificial Intelligence, digital security, etc.) into manufacturing tools, machines and robots, industrial processes, factories environment, supply chain integration, intra logistics & logistics 4.0 product life cycle, smart packaging, etc.
Application Deadline in a day

ESA's Call for Proposals: Space and Biotechnology

The European Space Agency’s “Space and Biotechnology” opportunity offers support and funding to companies looking to develop biotechnology services that can benefit from using space-based data or capabilities.
Application Deadline in a month

The Shell GameChanger™ Accelerator Powered by NREL Announces First Cohort of Start-Ups

Each participating startup will receive as much as $250,000 in non-dilutive funding to aid in development and demonstration of its technology, with the opportunity for future follow-on funding and beta-testing with a strategic program partner, in addition to access to expertise and laboratory capabilities at both NREL and Shell.