University Events

Description: Topic: Hyperparameter Optimization in Convolutional Neural Network Using Genetic Algorithm with Stopping Criterion

ZOOM Teleconference: https://umassd.zoom.us/j/98584029023

Abstract:
Hyperparameter optimization is the maximization of a neural network's accuracy using a set of input hyperparameters. With much research going into neural networks, in this case convolutional neural networks (CNNs), most research uses manually chosen hyperparameters instead of choosing optimal hyperparameters. Much work has gone into optimizing hyperparameters as they greatly affect the accuracy of CNNs and enable automated machine learning. Choosing the optimal hyperparameters requires applying an algorithm, in this case the genetic algorithm, to search for them. Each search operation requires a full training using a set of hyperparameters which is computationally expensive. New research has shown advantages to applying a stopping criterion to improve the neural network accuracy while reducing the computational cost. The goal of this research project is to use the genetic algorithm with adapted stopping criterion to reduce the computational cost of optimizing hyperparameters in CNNs.

NOTE: All ECE Graduate Students are ENCOURAGED to join the zoom teleconference. All interested parties are invited to join.

Advisor: Dr. Ruolin Zhou
Committee Members: Dr. Honggang Wang, and Dr. Liudong Xing, Department of Electrical & Computer Engineering, University of Massachusetts Dartmouth

*For further information, please contact Dr. Ruolin Zhou via email at rzhou1@umassd