This paper proposes Deep Hyperalignment (DHA) as a regularized, deep extension, scalable Hyperalignment (HA) method, which is well-suited for applying functional alignment to fMRI datasets with nonlinearity, high-dimensionality (broad ROI), and a large number of subjects. Unlink previous methods, DHA is not limited by a restricted fixed kernel function. Further, it uses a parametric approach, rank-m Singular Value Decomposition (SVD), and stochastic gradient descent for optimization. Therefore, DHA has a suitable time complexity for large datasets, and DHA does not require the training data when it computes the functional alignment for a new subject. Experimental studies on multi-subject fMRI analysis confirm that the DHA method achieves superior performance to other state-of-the-art HA algorithms.
Muhammad Yousefnezhad (Nanjing University of Aeronautics and Astronautics)
I am the Director of Brain Decoding section, iBRAIN group, Department of Computer Science & Technology, Nanjing University of Aeronautics and Astronautics. We are developing Artificial Intelligence algorithms in order to understand (decode) generated patterns in the human brain. Most of my counterparts try to change the world, but I am first trying to understand how it works!
Daoqiang Zhang (Nanjing University of Aeronautics and Astronautics)
Related Events (a corresponding poster, oral, or spotlight)
2017 Spotlight: Deep Hyperalignment »
Thu. Dec 7th 08:05 -- 08:10 PM Room Hall A
More from the Same Authors
2020 Poster: Shared Space Transfer Learning for analyzing multi-site fMRI data »
Tony Muhammad Yousefnezhad · Alessandro Selvitella · Daoqiang Zhang · Andrew Greenshaw · Russell Greiner