Location
https://ccse.kennesaw.edu/computing-showcase/cday-programs/spring2021program.php
Document Type
Event
Start Date
26-4-2021 5:00 PM
Description
Reducing the learning rate of a CNN can positively affect the validation accuracy of a machine learning model. Dropping out nodes from different layers can further delay overfitting from happening. Validation loss decreases over more epochs, but it must be cut when it reaches its minimum value.Advisors(s): Dr. Dan LoTopic(s): Artificial Intelligence
Included in
UR-65 CNN CIFAR Image Identification
https://ccse.kennesaw.edu/computing-showcase/cday-programs/spring2021program.php
Reducing the learning rate of a CNN can positively affect the validation accuracy of a machine learning model. Dropping out nodes from different layers can further delay overfitting from happening. Validation loss decreases over more epochs, but it must be cut when it reaches its minimum value.Advisors(s): Dr. Dan LoTopic(s): Artificial Intelligence