Yibin (Spencer) Sun

I am Yibin (Spencer) Sun, a machine learning researcher affiliated with Victoria University of Wellington and the University of Waikato. My work focuses on adaptive learning, data stream mining, and reliable real-time prediction.

About

Learning from data that never stops.

I hold a PhD in Computer Science from the University of Waikato in New Zealand, completed in 2025. I develop machine learning methods for streaming data, where models need to learn continuously as the world changes.

My broader interests include concept drift, online continual learning, uncertainty quantification, dynamic ensemble selection, and open-source tools that make stream learning accessible. I also apply these methods to energy pricing and electromyography (EMG) signals.

Research & open source

Current areas of contribution

01

Adaptive learning

Incremental models that respond to concept drift and learn continuously from non-stationary data.

02

Uncertainty quantification

Fully incremental prediction intervals and evaluation methods for reliable streaming regression.

03

Dynamic ensembles

Adaptive member selection that balances accuracy and diversity as classification streams evolve.

04

Applied stream learning

Real-time methods for New Zealand energy pricing and domain-shift detection in EMG signals.

05

CapyMOA

An open-source Python library for efficient data stream learning and online continual learning.

Visit CapyMOA

06

Massive Online Analysis

A Java-based open-source platform for building and evaluating data stream learning systems.

Visit MOA

Selected publications

Recent work

  1. 2026

    Adaptive approaches towards fully incremental prediction interval for data stream regression

    Sun, Y., Pfahringer, B., Gomes, H.M. et al. · Knowledge and Information Systems, 68, 227

    Read paper

  2. 2025

    Detecting Domain Shifts in Myoelectric Activations: Challenges and Opportunities in Stream Learning

    Sun, Y., Lim, N., Cassales, G.W., et al. · PRICAI 2025

    Read paper

  3. 2025

    Dynamic Ensemble Member Selection for Data Stream Classification

    Sun, Y., Pfahringer, B., Gomes, H.M., Bifet, A. · CIKM 2025

    Read paper

  4. 2025

    CapyMOA: Efficient Machine Learning for Data Streams and Online Continual Learning in Python

    Gomes, H.M., Lee, A., Gunasekara, N., Sun, Y., et al. · arXiv preprint

    Read paper

  5. 2025 · PhD thesis

    Improving ensembles and prediction intervals for machine learning on data streams

    Sun, Y. · University of Waikato

    Read thesis

  6. 2024

    Adaptive Prediction Interval for Data Stream Regression

    PAKDD 2024

    Read paper

  7. 2024

    Real-Time Energy Pricing in New Zealand: An Evolving Stream Analysis

    PRICAI 2024

    Read paper

  8. 2022

    SOKNL: Integrating K-nearest neighbours with adaptive random forest regression for data streams

    ECML PKDD 2022

    View on Scholar

Talks & tutorials

Sharing stream learning research

2024 · Tutorial

Machine Learning on Streaming Data

PRICAI and ICONIP

2024 · Presentation

Adaptive Prediction Interval for Data Stream Regression

New Zealand AI Summit

2022 · Presentation

Self-Optimising K Nearest Leaves

ECML PKDD

Contact

yibin.spencer.sun@gmail.com

Victoria University of Wellington · University of Waikato · New Zealand