Arman Kazemi's work published in Scientific Reports

Author: Michael Niemier

Src Logo Small

Work from recent group alumni Arman Kazemi (now at Amazon) and current group member Mehdi Sharifi has been accepted and published in Scientific Reports.  Arman's work addresses whether or not it is possible to achieve iso-accuracy with technology enabled architectures for hyperdimensional computing problems.  This work is important as, frequently, accuracy may be sacrificed as new device/architectural solutions are applied to machine learning/AI algorithms.  Arman and Mehdi's work is important as it suggests how one can maintain iso-accuracy with software-based solutions.  This work was been sponsored by the SRC/DARPA JUMP program.  The complete article can be found here.