Interactive Hyperparameter Optimization with Paintable Timelines
Authors
We propose a method to integrate more interactivity into automatic hyperparameter optimization systems to leverage the user's prior knowledge on parameter distribution. In our method, the user continuously observes automatic optimization's progress and dynamically specifies where to search in the parameter space. We present a prototype implementation of an interactive dashboard for an optimizer to show our method's feasibility. The interactive dashboard's main feature is ``paintable timeline'' where the user can not only observe the past parameter values tested as in standard timeline but also specify the range of future parameters to be tested with simple painting operations. We show three examples where user intervention might improve the performance of automatic optimizations. We run a user study with experts and the results show that, with prior knowledge about parameter distribution of the target problem, interactive optimization can reach better results compared to fully automatic optimization.
Research Questions / Practical Problems
Question signals indexed for this paper.
- 100%
Model Compression in Practice: Lessons Learned from Practitioners Creating On-device Machine Learning Experiences
CHI '24· Human-LLM Collaboration +1
- 100%
InstructPipe: Generating Visual Blocks Pipelines with Human Instructions and LLMs
CHI '25· Human-LLM Collaboration +1
- 100%
Text-to-SQL Domain Adaptation via Human-LLM Collaborative Data Annotation
IUI '25· Human-LLM Collaboration +1
- 80%
AI for Low-Code for AI
IUI '24· Generative AI (Text, Image, Music, Video) +2
- 80%
Genie in the Model: Automatic Generation of Human-in-the-Loop Deep Neural Networks for Mobile Applications
UbiComp '23· Human-LLM Collaboration +2
- 67%
Towards Human-Guided Machine Learning
IUI '19· Human-LLM Collaboration +2
- 67%
CoAutoML: User Interface Framework for Machine Learning Novices using LLM-based AutoML and Test-Driven Machine Teaching
IUI '26· AutoML Interfaces +2
- 67%
Never-ending Learning of User Interfaces
UIST '23· Human-LLM Collaboration +2
- 60%
Trade-offs for Substituting a Human with an Agent in a Pair Programming Context: The Good, the Bad, and the Ugly
CHI '21· Human-LLM Collaboration +1
- 60%
Visualizing Examples of Deep Neural Networks at Scale
CHI '21· Human-LLM Collaboration +1
Based on Jaccard similarity of research subtopics & professions (≥60%)