Student Posters Repository

Recommendation systems become a much needed tool for predicting user preferences. The development of recommender systems subsume in addition to information retrieval, models and metrics creating, detection of key parameters and many machine learning applications. Model creating is a computationally demanding problem, that can use heterogeneous models as a tool for more effective computational distribution. The poster presents a combination of heterogeneous models and the problem of preference recommendation by using matrix factorization.

There is a rich body of numerical, experimental and theoretical work looking at the role of turbulence in particle collisions, with a particular emphasis on how it might accelerate the evolution of clouds in the atmosphere. This study is a continuation of that lineage. I perform direct numerical simulations of isotropic turbulence with embedded droplets that, upon collision, coalesce to produce a daughter droplet that conserves the mass and momentum of the parent droplets. As a consequence of coalescence, the droplet size distribution evolves over time from its monodisperse initial condition. The work is an extension of Reade and Collins (J. Fluid Mech. 415:45-64, 2000), which considered the same problem at a much lower Reynolds number. Important effects of intermittency at Reynolds numbers that are several-fold higher are observed. My simulation will take into account those intermittency effects and consequent influence on the evolution of the size distribution with varying particle size, volume fraction and number density. The collisions do not yet take into account the effect of the lubricating gas layer, which will be the topic of my future work 

Real time human face processing is a quite challenge in today's scientific study. How to save the time and let the system work much more rapid is the core target to this problem.

It is comprehensive work in different areas such as network, data transmission, cloud computing, objective detection and recognition, etc. Our main study is to use the current stable and efficient algorithm and make it in parallel version to find a faster way to improve this progress. In this year, we have tested much bigger database and make our algorithm much more stable than before.