Sandeep Santosh

About

Sandeep Santosh is a researcher at the forefront of efficient, real-time decision-making systems, specializing in the hardware acceleration of Multi-Armed Bandit (MAB) algorithms. His work addresses a critical challenge: deploying these powerful algorithms—which balance exploration and exploitation to identify optimal choices in unknown environments—onto resource-constrained edge devices for applications in wireless radio, IoT, and robotics. Santosh’s major contributions include the development of intelligent, reconfigurable architectures for MAB algorithms, notably a KL divergence-based design and a computationally efficient framework, both published in 2020. He has also provided a critical comparative analysis of Frequentist versus Bayesian approaches on Zynq System-on-Chip platforms (2022). With his most-cited papers accumulating over 37 citations, Santosh’s work is establishing foundational hardware solutions for autonomous systems. His research extends to practical automation, as seen in his work on machine learning-driven robotic arms, demonstrating a clear commitment to bridging algorithmic theory with tangible, deployable hardware.

Research Focus

Key Achievements

Top Papers

1. 1
   Intelligent and Reconfigurable Architecture for KL Divergence-Based Multi-Armed Bandit Algorithms

   12 citations · 2020
2. 2
   Multiarmed Bandit Algorithms on Zynq System-on-Chip: Go Frequentist or Bayesian?

   11 citations · 2022
3. 3
   An Automated Robotic Arm: A Machine Learning Approach

   9 citations · 2021
4. 4
   Reconfigurable and Computationally Efficient Architecture for Multi-Armed Bandit Algorithms

   5 citations · 2020

Key Collaborators

Contact & Links