Pragya Sharma

Pragya Sharma

PhD, School of Electrical and Computer Engineering, Cornell University, Prof. Edwin Kan Group

Accuracy and Resources Trade-off in Machine Learning Algorithms

Cornell University

2019

This course project involved approximation techniques for resource-constrained scenarios

  • Developed approximation techniques to tackle constrained-resource issues with large training data and limited CPU, memory, and energy availability, specifically in mobile devices.
  • Studied accuracy vs resource trade-off with different approximation techniques at both training and inference in machine learning algorithms including regression, neural networks, and support vector machine (SVM).