Senior Scientist Head of Cognitive Security Technologies Fraunhofer Institute for Applied and Integrated Security Address Fraunhofer AISEC Institute Parkring 4 85748 Garching bei München, Deutschland Tel.: +49 89 32299 86 149 Fax.: +49 89 32299 86 299 Emails Office hours By appointment

The world is in the process of digitalization, human is facing revolutions through the golden age of modern years. We are pondering of what's changed and what's staying immortal. The cognitive level of non-human objects is increasing to a danger frontier or not? we question ourselves, challenge others, and pray for the future.

Recent posts
Notes of experimenting Holmes-Processing system implemented by my colleague George Webster. It meant to be a large-scale distributed framework for security analysis. ...more
For writting LaTeX, it's normal that you will have to install a third party library to enhace your work. This post shows you how to install a third package in your MikTeX distribution. ...more
We'll review a paper regarding botnets detection. An application on botnets detection will give us an idea of what and how we can do about security problems leveraging machine learning algorithms. ...more
This tutorial aims at those who have multiple github accounts and wanna maintain them on local machines. We walk you through how you can setup multiple ssh keys for different accounts. ...more
Here I list a set of Machine Learning questions for interview who wants to get a data scientist job or similar position. Answers can be inquired by contact. ...more
ICML conference is well organized in terms of main events, incl. presentations, posters, tutorials and workshops. However catering and services are kinda of being compromised due to a huge amount of visitors, ~1600 ...more
Selected Projects
We open sourced a grammatical inference toolbox hosted on Google code. The toolbox is implemented in Matlab, and contains several GI algorithms. ...more
We use bokeh to visualize some common plots in Machine Learning, and we make some wrappers to make it even easier to use particularly for research. ...more
Support vector clustering is a kernel-based unsupervised learning algorithm used mainly for anomaly deteciton. Here I give a detailed tutorial on understanding this type of algorithm. ...more