When two nuclei grew and the distance between them was reduced, a liquid channel developed and facilitated their coalescence.
due to that the energy of the solid/liquid interface is lower than that between solid phases.
Currently, I am a Postdoctoral Fellow working on simulation and theoretical modeling in soft matter physics and statistical mechanics in Prof. Yilong Han's group, in the Department of Physics, Hong Kong University of Science and Technology.
I am a big fan of physics and math, especially statistical physics and complex systems. I'm fascinated by the rich collective behaviors of systems governed by simple rules of microscopic dynamics. Those help me to understand the nature and even the society. I'm interested in all aspects of statistical physics, such as phase transitions, kinetics, non-equilibrium processes, complex networks, and other many that I haven't deepen into, e.g. evolutionary dynamics, collective motions/patterns in biological systems. I prefer theoretical work and simulation than experiments, since I believe this can help to reach to more fundamental understandings and make experimental data meaningful.
I also like history, and history-based video games, such as the Total War series, Mount & Blade. I am interested in how an empire thrives and collapses, and how to command a huge military troop in a battle field such as besieging a city or fighting in mountains.
Phase transitions happens in our daily life: melting of ice, boiling of water, etc. Equilibrium statistical mechanics teaches us when a phase transition may happen and the initial and finial phases of the system, however, it tells nothing on how the transition happens, especially for first order transitions. Understanding of those transition kinetics has not only great important applications in industry, but also helps to obtain a complete theory on many-body systems.
Currently, I am working on the crystal melting process in Prof. Yilong Han's group. I conduct theoretical calculations as well as event-driven molecular dynamics (EDMD) simulations to address the question on how a hard sphere crystal melts and when a superheated crystal melts immediately without following a nucleation pathway.
The most famous study on the microbiome is the celebrated Human Microbiome Project (HMP). The collected data is open and can be found on its official websites.
1011 (100 billion) microbes are living in our human body, and most of them are in our gut. This number is 10 times more than human cells, and more importantly, the gene diversity is about 100 times than human. Those microbes form huge ecological communities, perform very complex metabolic functions and interact with our human cells. Recently developed next-generation sequencing technique enables us to reveal both the members of this huge community as well as its functions and impact on our human health. For example, it is found that the microbes are related or contributing to obesity, Type-1 diabetes, inflammatory bowel disease, some mental diseases, and so on.
I am interested in applying ecological models (e.g. neutral models, GLV models) and machine learning algorithms into the microbial community, to reveal its temporal dynamics, ecological species-species interaction, as well as their impact on human health, infant/child development, and aging of old people.
A matrix with all elements are independent identical random numbers follows a universal law -- Girko circular law (1984) -- the eigenvalues of this kind of matrices are distributed uniformly over a disk on complex plane.
Imposing constraints can not only change the uniformness of eigenvalue distribution, but also the shape and topology of the distribution. For example, eigenvalue distribution of 106 randomly generated Sudoku 9×9 matrix shows a band gap structure on complex plane.
In application, those eigenvalue distributions are related to energy bands of disordered systems, phase transitions of disordered systems, stability of complex ecological system, etc. It could also be viewed as a mathematical art.
I am interested in exploring the eigenvalue distributions of various kinds of matrices, and trying to figure out the universal features of the distributions. It is also important to explore the localization of the corresponding eigenvectors, which is associated with the responds, information, or phase transition of the systems in applications.
Here is a gallery of the eigenvalue distributions of some interesting matrices:
http://nbviewer.jupyter.org/url/jakevdp.github.com/downloads/notebooks/SudokuCodeGolf.ipynb provides a one-sentence Python function to solve Sudoku.
Given a list of references (e.g. zotero research papers), and a list of keywords, we might want to ask, what is the relationship between those keywords among those research papers.
Here we use network concept:
node: keyword
node weight: number of research papers containing the keyword
edge: research papers containing both keywords
edge weight: number of research papers containing both keywords
The resulted undirected weighted network represents the keyword association among the papers in the database. This may be useful if one would like to know how the overall concepts in one's research area look like. An example is shown as the right figure.
The network can be further visualized interactively by d3.js so that one can obtain the whole paper lists by putting mouse on the node or edge. The interactive network can be found here, and the source code and documentation are on Github.
My name, Feng Wang, is a very common name in Chinese, and also very common in academia. Searching my name with google, one can find a famous singer, several professors in University of California Berkeley (physics), University of Hong Kong (physics), University of Michigan (social science), etc. on the top page of the google's searching results.
It seems that my name is beyond the top 50 of common Chinese names given in 2007.
Anyway, it is almost impossible to find me by a google search. Frustrated...