Rakuten Institute of Technology San Mateo’s research scientist Ramin Raziperchikolaei attended the 36th International Conference on Machine Learning (ICML). ICML is dedicated to presenting and publishing research on machine learning which includes artificial intelligence, statistsics and data science, as well as robotics, computational biology, and machine vision. ICML was held from June 9-June 15 at the Long Beach Convention & Entertainment Center in Long Beach, California with over 6,000 participants in attendance.
Ramin's collaborative research paper with advisor Harish Bhat of UC Merced, “A Block Coordinate Descent Proximal Method for Simultaneous Filtering and Parameter Estimation,” analyzes a block coordinate descent proximal algorithim (BCD-prox) to solve filtering and parameter estimation of ordinary differential equations (ODE). In so doing, they demonstrate that BCD-prox exhibits improved accuracy, increased robustness, and decreased training times.
"As a machine learning scientist, there is nothing more enjoyable than attending ICML and presenting your work. By attending ICML, I discussed ML topics with other researchers in different areas, attended great talks, and came up with new ideas for my current research at Rakuten."
As we continue to conduct research, we hope that we will be able to participate and make impactful contributions at ICML in the future.