Project: Benchmarking a Multimodal Time Series Dataset
Description
1. Introduction
Multimodal time series datasets are increasingly valuable in finance, healthcare, industrial monitoring, and other domains. However, their availability remains limited, and standardized benchmarking is underexplored. This project benchmarks a new multimodal time series dataset from the company WAIR, assessing its unique characteristics and positioning it relative to existing public datasets. The study aims to ensure the dataset’s readiness for public release and contribute to research in multimodal time series analysis.
2. Research Objectives
This study aims to benchmark the multimodal time series dataset against existing ones, explore its statistical properties, conduct experimental evaluations, define dataset quality metrics, and provide insights for its publication.
3. Expected Contributions
(1) You will first need to do a literature review on existing multimodal datasets, maybe not limited to time series. (2) You would need to do a deeper analysis on the multimodal dataset, designing and conducting different experiments to evaluate its different properties, diverse patterns, and qualities, good or bad. (3) You are also expected to involve time series predictive models to assess the performance of a prediction model using the multimodal datasets. (4) You would also possibly consider out-of-distribution generalization problems. (5) You would be able to put experimental studies into your paper writing.
4. Goal
This project will provide a structured benchmarking study, ensuring the dataset’s readiness for public release and positioning it as a standard for future research. Through comparative analysis and experimental evaluation, it will highlight the dataset’s strengths and limitations.
- Paper publication is highly expected.
- Experiences in time series projects are high favourable.
You will be mainly supervised by Dr. Deng for the whole process, you will also be able to collaborate with the company WAIR.
Details
- Student
-
ZTZhang Tianyi
- Supervisor
-
Amy Deng