Computer Science > Human-Computer Interaction
[Submitted on 24 May 2021]
Title:Guided Hyperparameter Tuning Through Visualization and Inference
View PDFAbstract:For deep learning practitioners, hyperparameter tuning for optimizing model performance can be a computationally expensive task. Though visualization can help practitioners relate hyperparameter settings to overall model performance, significant manual inspection is still required to guide the hyperparameter settings in the next batch of experiments. In response, we present a streamlined visualization system enabling deep learning practitioners to more efficiently explore, tune, and optimize hyperparameters in a batch of experiments. A key idea is to directly suggest more optimal hyperparameter values using a predictive mechanism. We then integrate this mechanism with current visualization practices for deep learning. Moreover, an analysis on the variance in a selected performance metric in the context of the model hyperparameters shows the impact that certain hyperparameters have on the performance metric. We evaluate the tool with a user study on deep learning model builders, finding that our participants have little issue adopting the tool and working with it as part of their workflow.
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.