Reseach Projects

Swarm Intelligence

We have done a lot of work on novel approaches in the filed of swarm intelligence. Recently, we proposed Generative Adeversarial Optimization (GAO), which combines neural network and black-box optimization. And we proposed Fireworks Algorithm (FWA) in 2010 which is a powerful swarm intelligence algorithm and attracted lots of attention. Historically, we also studied Particle Swarm Optimization. More...

1. Fireworks Algorithm (FWA)

Fireworks Algorithm is a recently developed swarm intelligence algorithm.. More...

2. Generative Adversarial Optimization (GAO)

A optimization technique based on Generative Adeversarial training. More...[code]

4. Cooperative Multi-Agent Reinforcement Learning(MARL)

We have done some mutl-agent reinforcement learning research. More...

5. Group Robots Multiple Targets Search

Simple cooperative rules can lead to complex behaviors when applied to a group of robots. More ...


Machine Learning

We also have accumulated lots of novel work in the filed of machine learning. In recent years, we've focused on applying modern deep learning techniques to financial markets and natural language processing.

5. AiFinance

We've done some research on how to apply deep neural networks to stock market risk prediction and stock ranking. We've also developed our simulated trading platform and Wechat mini program. More ...

6. Natural Language Processing

Text classification methods based on semi-supervised learning and variational autoencoder. More...

7. Learning based Anti-Spam Techniques

Spam filtering plays an important role in modern society. We have devleloped MalGAN (Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN) and uninterrupted detection approaches based on Incremental Support Vector Machine and Artificial Immune System for the spam of e-mail stream. More...


Historical Research

8. HEV control strategy research based on the computational intelligence

We have proposed an efficient approach for forecasting the driving load of the hybrid electric vehicle (HEV). More...

9. Immune based computer malware detection

We have proposed several malware feature extraction methods and malware detection models based on immune theory. More...

10. Genetic Programming Based Morphological Image Analysis Algorithm

We have proposed an Genetic Programming Based Morphological Image Analysis Algorithm. More...

11. FRT

A deinition of the frequency of an image is given. And a frequency transform (FRT) based on the definition is developed. More...


Scientific Grants

  • The National High Technology Research and Development Program of China (863 Program), with Grant No. 2007AA01Z453, (2008-2012).
  • National Natural Science Foundation of China (NSFC), with Grants No. 60673020, (2007-2009).

International Collaborations

  • The Chinese University of Hong Kong, Forecast of Drive Load by Computational Intelligence.
  • Stevens Institute of Technology, USA. Spam Detection by Machine Learning Method.

Technical Reports

Copyright © 2007-2012 Computational Intelligence Laboratory, Peking University. All Rights Reserved
Tel: 86-10-62756374 PostCode: 100871
Address: School of Electronics Engineering and Computer Science, Peking University