Scientists have theories of how humans learn concepts from a robot’s perspective. MIT and Harvard researchers apply these theories to challenges in human-robot interaction. The results show how the theories can help humans form conceptual models of robots more quickly, accurately, and flexibly.
Source: https://neurosciencenews.com/human-robot-understanding-20125/
The field of machine learning on quantum computers got a boost from new research removing a potential roadblock to the practical implementation of quantum neural networks. While theorists had previously believed an exponentially large training set would be required to train a quantum neural network, the quantum No-Free-Lunch theorem developed by Los Alamos National Laboratory shows that quantum entanglement eliminates this exponential overhead. A direct consequence of this theorem that showcases the power of data in classical machine learning is that the more data one has, the better the average performance.
Source: https://phys.org/news/2022-02-entanglement-scaling-quantum-machine.html
A physicist at the University of Michigan is utilizing quantum computing and algorithms to better comprehend the concept of holographic duality. Holographic duality is a mathematical hypothesis that links particle theories and dynamics with gravity theory. This hypothesis implies that the theories of gravity and particles are numerically equal: whatever occurs in the theory of gravity also arises in the theory of particles and vice versa. Our whole world, according to some scientists, is a holographic projection of particles, which might lead to a consistent quantum explanation of gravity.
Source: https://phys.org/news/2022-02-black-hole-physicist-quantum-machine.html
From the autumn of 2022, Stanford University’s Machine Learning with Graphs course will be offered online for free. The course focuses on the computational, algorithmic, and modeling issues that come with large-scale network analysis. Students are taught machine learning techniques and data mining tools capable of revealing insights on a range of networks by investigating the underlying graph structure and its properties. Representation learning and Graph Neural Networks are among the subjects taught in the course, as are algorithms for the World Wide Web, reasoning over Knowledge Graphs, impact maximization, disease outbreak detection, and social network analysis.
Ghassemi and her collaborator, Boston University’s Elaine Okanyene Nsoesie, wrote a warning note on the future of AI in medicine in a piece published on Jan. 14 in the journal Patterns. According to Ghassemi, if utilized properly, technology may enhance healthcare performance and perhaps eliminate disparities, but if we aren’t vigilant, it might really impair treatment. Given that the AI systems in issue train themselves by processing and analyzing massive amounts of data, it all boils down to data. The goal of the research is not to prevent engineers from applying their machine learning knowledge in the medical field.
Source: https://news.mit.edu/2022/marzyeh-ghassemi-explores-downside-machine-learning-health-care-0201
