Basic Theories, Algorithms, and Applications Related to Bayesian Statistics, Machine Learning (especially Deep Learning), Optimization Methods, and Signal Processing

Dr. LIU has led and completed some research projects, including 1 general program and 1 young scholar program from the National Natural Science Foundation of China, 1 first-class funding project and 1 special funding project from the Chinese Postdoctoral Science Foundation, 2 research projects from Natural Science Foundations of the provincial and ministerial level, 1 international exchange program of the Royal Society of UK, 1 open project from a State Key Lab and 3 enterprise R&D projects. Currently, as the principal investigator, Dr. LIU is leading a team at Zhejiang Lab, researching basic theories and algorithms related to deep learning and Bayesian inference.

For Dr. LIU’s paper publications, see：

Google Scholar: https://scholar.google.com/citations?user=UWCOeegAAAAJ&hl=en

Researchgate: https://www.researchgate.net/profile/Bin-Liu-120

Representative Research Achievements Include:

1. The theory of Bayesian dynamic ensemble of multiple models (BDEMM), and a series of algorithms developed based on it, including non-stationary neural decoding algorithm for brain-computer interfaces; robust particle filtering; robust sequential online prediction method that is immune of outliers and change points; accelerated Bayesian optimization method based on effective leveraging of low-fidelity data.

2. A label-free model selection method for convolutional neural networks based on the principle of consistent relative confidence.

3. Global optimization algorithm based on sequential Monte Carlo methods.

4. Adaptive annealed importance sampling (AAIS) algorithm, and an AAIS based approach to extrasolar planet detection.

5. A Dirichlet Process Mixtures (DPM) and Gaussian Process (GP) based electromagnetic response surface fitting algorithm for Rapid design of metamaterial microstructures.

6. A wireless communication MIMO decoding algorithm based on matrix decomposition and Markov Chain Monte Carlo sampling.

7. The theory of Generic Bayesian Trust (GBT), and GBT based robust data analysis methods for sensor networks.

Our team focuses on innovative and technological research of basic theories and algorithms related to machine learning and Bayesian inference. The team members include research fellows (including postdocs) who focus on basic theories and algorithms, algorithm engineers specialized in machine learning or high-performance computing. Our team has adequate hardware and software facilities, a strong atmosphere of innovation, and a pure technology-oriented team culture. We are looking for talents to join us (if you are interested, just send an e-mail to Liubin AT zhejianglab.com).

Jobs:

1. Job Description and Application Link for Postdoc Positions:

https://zp.zhejianglab.com/detail.aspx?id=F803E4F6E2744A0DB020656CFB3159DB

2. Deep Learning Research Fellow：

Job Description and Application Link for Social Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=BAE3AA0D28BA48EFB8FDE81026C21298

Job Description and Application Link for Campus Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=478BE73B258046E19FC679B452308D4F

3. Deep Learning Algorithm Engineer：

Job Description and Application Link for Social Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=26B3945B3ECE43F4949412876CA4A8FD

Job Description and Application Link for Campus Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=52B4B420624E41DF8E08F88762B2F514

4. High-Performance Computing Algorithm Engineer：

Job Description and Application Link for Social Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=F92B8EA921664537AED916C922F53E15

Job Description and Application Link for Campus Recruitment:

https://zp.zhejianglab.com/detail.aspx?id=FB1BB26DE5CC4CE891A6F350D90094D3