Software/Code

This page contains links to software and codes developed as part of my and my students' research. 

Sparse Graph Learning in Function to Function Regression  with Graph-induced Operator-valued Kernels 

Code for the TMLR 2024 paper Learning Sparse Graphs for Functional Regression using Graph-induced Operator-valued Kernels by Akash Saha, P. Balamurugan is available at the following link: [code link]

Distributed Accelerated Gradient Methods with Restart under Quadratic Growth Condition

Code for the JOGO 2024 paper Distributed Accelerated Gradient Methods with Restart under Quadratic Growth Condition by Chhavi Sharma, Vishnu Narayanan, P. Balamurugan is available at the following link: [code link]

Switch and Conquer: switching scheme between SGD and SVRG oracles for decentralized saddle-point problem optimization       

Code for the IEEE CDC 2023 paper Switch and Conquer: Efficient Algorithms By Switching Stochastic Gradient Oracles For Decentralized Saddle Point Problems by Chhavi Sharma, Vishnu Narayanan, P. Balamurugan, is available at:  [code link] 

Operator Minimum Residual (OpMINRES) Algorithm for Learning with Operator-valued Kernels in Reproducing Kernel Krein Spaces

Code for the NeurIPS 2020 paper Learning with Operator-valued Kernels in Reproducing Kernel Krein Spaces by Akash Saha, P. Balamurugan, is available at:  [code link] 

Stochastic variance reduced gradient method for single-machine saddle-point optimization problem

Code for the NIPS (now NeurIPS) 2016 paper Stochastic Variance Reduction Methods for Saddle-Point Problems by P. Balamurugan, Francis Bach, is available at:  [code link]

Sequential Dual Algorithms for Structural Support Vector Machine

Codes for sequential dual method for ℓ2 regularized structural SVM,  sequential alternating proximal method for elastic net regularized structural SVM and an ADMM method for ℓ1+ℓ2 regularized structural SVM problem are available at: [code link]