Researchers have developed a new framework for deep neural networks that allows artificial intelligence (AI) systems to better learn new tasks while “forgetting” less of what they have learned regarding previous tasks.
The researchers have also demonstrated that using the framework to learn a new task can make the AI better at performing previous tasks, a phenomenon called backward transfer.
“People are capable of continual learning; we learn new tasks all the time, without forgetting what we already know,” says Tianfu Wu, an assistant professor of electrical and computer engineering at NC State and co-author of a paper on the work. “To date, AI systems using deep neural networks have not been very good at this.”
“Deep neural network AI systems are designed for learning narrow tasks,” says Xilai Li, a co-lead author of the paper and a Ph.D. candidate at NC State. “As a result, one of several things can happen when learning new tasks. Systems can forget old tasks when learning new ones, which is called catastrophic forgetting. Systems can forget some of the things they knew about old tasks, while not learning to do new ones as well. Or systems can fix old tasks in place while adding new tasks – which limits improvement and quickly leads to an AI system that is too large to operate efficiently. Continual learning, also called lifelong-learning or learning-to-learn, is trying to address the issue.”
“We have proposed a new framework for continual learning, which decouples network structure learning and model parameter learning,” says Yingbo Zhou, co-lead author of the paper and a research scientist at Salesforce Research. “We call it the Learn to Grow framework. In experimental testing, we’ve found that it outperforms previous approaches to continual learning.”
To understand the Learn to Grow framework, think of deep neural networks as a pipe filled with multiple layers. Raw data goes into the top of the pipe, and task outputs come out the bottom. Every “layer” in the pipe is a computation that manipulates the data in order to help the network accomplish its task, such as identifying objects in a digital image. There are multiple ways of arranging the layers in the pipe, which correspond to different “architectures” of the network.
When asking a deep neural network to learn a new task, the Learn to Grow framework begins by conducting something called an explicit neural architecture optimization via search. What this means is that as the network comes to each layer in its system, it can decide to do one of four things: skip the layer; use the layer in the same way that previous tasks used it; attach a lightweight adapter to the layer, which modifies it slightly; or create an entirely new layer.
This architecture optimization effectively lays out the best topology, or series of layers, needed to accomplish the new task. Once this is complete, the network uses the new topology to train itself on how to accomplish the task – just like any other deep learning AI system.
“We’ve run experiments using several datasets, and what we’ve found is that the more similar a new task is to previous tasks, the more overlap there is in terms of the existing layers that are kept to perform the new task,” Li says. “What is more interesting is that, with the optimized – or “learned” topology – a network trained to perform new tasks forgets very little of what it needed to perform the older tasks, even if the older tasks were not similar.”
The researchers also ran experiments comparing the Learn to Grow framework’s ability to learn new tasks to several other continual learning methods, and found that the Learn to Grow framework had better accuracy when completing new tasks.
To test how much each network may have forgotten when learning the new task, the researchers then tested each system’s accuracy at performing the older tasks – and the Learn to Grow framework again outperformed the other networks.
“In some cases, the Learn to Grow framework actually got better at performing the old tasks,” says Caiming Xiong, the research director of Salesforce Research and a co-author of the work. “This is called backward transfer, and occurs when you find that learning a new task makes you better at an old task. We see this in people all the time; not so much with AI.”
The Latest on: Continual learning for artificial intelligence
via Google News
The Latest on: Continual learning for artificial intelligence
- Designing biological-inspired artificial intelligenceon October 5, 2020 at 9:57 am
For decades artificial intelligence (AI) researchers have turned to the brain for inspiration but the mechanisms by which neurons create or suppress ...
- Visualizing defects in crystal solids revolutionized by Artificial Intelligenceon October 2, 2020 at 2:22 am
A new study challenges the ways in which defects in crystal structures are described, proposing a novel approach to characterizing and modeling crystal defects at the atomic scale based on machine ...
- Data Science to Expedite Drug Discovery with Artificial Intelligence and Machine Learning, Says Frost & Sullivanon September 29, 2020 at 5:00 pm
Data Science Impacting the Pharmaceutical Industry is the latest addition to Frost & Sullivan's Technical Insights research and analyses available via the Frost & Sullivan Leadership Council, which ...
- Data Science to Accelerate Drug Discovery with Artificial Intelligence and Machine Learning, Says Frost & Sullivanon September 29, 2020 at 8:40 am
Frost & Sullivan's recent analysis, Data Science Impacting the Pharmaceutical Industry, finds that data science tools are promising technologies transforming drug discovery costs, speed, and ...
- Rokt Recognized for Innovation in Artificial Intelligence with 2020 MarTech Breakthrough Awardon September 23, 2020 at 6:00 am
PRNewswire/ -- Rokt, the global leader in e-commerce marketing technology, today announced that it has been selected as the winner of the “Best Use of AI in MarTech” award in the third annual MarTech ...
- Artificial Intelligence Is Ready For Prime Time, But Needs Full Executive Supporton September 22, 2020 at 5:00 pm
Such is the case with eBay, the mega commerce site, that has been employing artificial intelligence ... Most importantly, AI is an iterative, continuous learning process, requiring automated ...
- DOE funding boosts artificial intelligence research at Jefferson Labon September 22, 2020 at 9:14 am
"Artificial Intelligence and machine learning have the potential to transform ... Their projects aim to improve operations of Jefferson Lab's Continuous Electron Beam Accelerator Facility, a ...
- Continuous Intelligence: The Newest Trend in Data Analysison September 22, 2020 at 6:32 am
Continuous intelligence (CI ... CI can leverage artificial intelligence (AI) and machine learning algorithms from a more analytical perspective. What makes CI such a compelling value proposition ...
- Kroger enlists artificial intelligence to cut down self-checkout errorson September 21, 2020 at 5:00 pm
Ireland-based Everseen said its artificial intelligence and machine learning platform began deployment ... of its stores “We are laser-focused on continuous improvements to customers ...
- K–12 IT Teams Use Artificial Intelligence to Help with Wi-Fi Upkeepon September 14, 2020 at 5:00 pm
The marriage of machine learning and Wi-Fi may help IT teams handle connectivity ... K–12 districts are adopting artificial intelligence to help with Wi-Fi upkeep as networks are feeling added ...
via Bing News