Artificial Intelligence Algorithm
The exponential growth of scientific papers in the field of artificial intelligence has made it practically difficult for human researchers to keep up with the latest developments in the field.
In a recent breakthrough, an international team of scientists led by Mario Krenn of the Max-Planck Institute for the Science of Light has created an artificial intelligence algorithm that does more than just help researchers find their bearings; it also provides predictive guidance as to the future trajectory of their own field of study. The study was published in Nature’s Machine Learning journal.
The number of research papers published on topics related to AI and ML doubles roughly every 23 months and is expanding at an exponential rate. It is extremely difficult for human researchers to keep up with the pace of development while still keeping a birds-eye vision of the big picture.
Read the latest blogs: Navigating The Reality Spectrum: Understanding VR, AR, and MR
The trajectory of Artificial Intelligence Study
The Max-Planck Institute for the Science of Light in Erlangen is led by Mario Krenn, who takes an unorthodox approach to solving this problem. Science4Cast, a new graph-based application he designed, allows users to ask questions regarding the trajectory of artificial intelligence study.
The multinational research team had previously launched the Science4Cast competition to capture and anticipate the growth of scientific concepts in the field of AI research, therefore identifying which themes would be the focus of future study. There were more than fifty submissions, all with unique perspectives.
Read special blogs: What Are B2B Robo-Advisors?
Krenn, along with the best teams, has analyzed the methodologies used, finding unexpected outcomes that span the spectrum from simply statistics to purely learning approaches. “The most effective methods use a carefully curated set of network features and not a continuous AI approach,” Mario Krenn stated. This hints at a sizable opportunity that can be tapped by purely ML methods independent of human expertise.
Read: AI and Machine Learning Are Changing Business Forever
What Will Happen?
As more and more scientific publications are published, Science4Cast’s graph-based representation of knowledge grows in complexity. Each node in the graph represents a different AI topic, and the edges linking them show whether or not the concepts were studied together and when.
The question “What will happen?” is a mathematical query regarding the graph’s future evolution, for instance. Over the course of 30 years, Science4Cast has been fed genuine data from over 100,000 scientific papers, yielding 64,000 nodes.
Predicting what scientists will study in the future is only the beginning. The authors of this study detail how future improvements to Science4Cast might one day provide academics with individualized recommendations for their next lines of inquiry.
[To share your insights with us, please write to sghosh@martechseries.com]
Comments are closed.