CMS Pixel Detector Upgrade

CMS Pixel Detector Upgrade

Mentor(s): Neeti Parashar

Undergraduate

Oral Presentation

: CMS (Compact Muon Solenoid) is a particle detector that is designed to see a wide range of particles and phenomena produced in high-energy collisions in the LHC (Large Hadron Collider). The Pixel Detector is designed to detect all the short-live particles. Higher energy of LHC will bring higher cross sections which current detector cannot perform with good efficiencies. There also are large hit rate in pixel detector and high data transfer requirements need to be taken care of. Those challenges push people to design a better pixel detector. Better performance and higher radiation hardness pixel detector will be necessary for the high energy physics research. This presentation will discuss the above mentioned issues and studies perform.

Generalization of the Weibull Distribution and its Applications

Mentor(s): Gokarna Aryal

Undergraduate

Oral Presentation

Probability distributions are important in all types of models and data analysis. The quality of statistical analysis is heavily dependent on the selection of probability distributions. Depending upon the assumptions made, the statistical interpretations can be very misleading. Thus, we look to build better models and distributions and have clear assumptions in order to make correct and better decisions.

Distribution fitting allows us to develop valid models of random processes which we deal with, protecting from potential time and money loss which can arise due to invalid model selection.

The Weibull distribution is a popular distribution in areas such as survival analysis, weather models, finance, and many other areas. This is due to the hazard rate function creating a “bathtub” shape, which can be used to model many real life phenomenon. The problem tends to be that the data rarely follows the standard Weibull distribution. Hence, we have looked to generalizing the Weibull distribution in order to have a distribution that better fits some data that we find in the real world.

In this study, we propose some methods of generalizing the Weibull distribution in order to capture the underlying behavior of data arising from several different areas. In particular, data related to survival times of cancer patients will be analyzed. We will also present the mathematical characteristics, method of estimating the parameters, and compare the goodness-of-fit of the models.

Geopolymer Concrete

Ivan Diaz

Mentor(s): Dr. Chen

Undergraduate

Oral Presentation

Geopolymer concrete is newly developed concrete that does not include Portland cement in the mix. This concrete can be used as an alternative while still maintaining the strength of traditional cement. The advantages of geopolymer concrete include reduction of use of waste material and use fewer harmful alkaline agents creating a green concrete.

Fly ash and slag are waste materials that come from refining metals and burning coal in power plants. These are waste materials that are often hazardous to our environment. By using geopolymer concrete waste material is being reduced by all parties involved.

Another advantage of geopolymer concrete in this project is formed using superplasticizer and sodium silicate mixed together in water. Both chemicals used can be safely handled with proper care without any harmful effects. This creates a more user-friendly concrete as opposed to traditional Portland cement.

The use of geopolymer concrete in the field can create a more sustainable environment. Reducing carbon dioxide emissions, hazardous chemicals being used, and waste products human expansion can be more viewed more environment friendly.

NFL Statistical Analysis Using R

Michael Tinich

Mentor(s): Gokarna Aryal

Undergraduate

Oral Presentation

This research which has been performed over the past two semesters is based on the National Football League (NFL) and the statistics that lie within the game of football. The primary propose of this study was to answer “how can we use the data gather during a game of football in a meaningful way to make right decisions and better predictions?” In my research we used a statistical programming language, R, to help sort large databases that included cumulative statistics and play by play statistics from every game in 2014 season. The dataset we used for our study are obtained from the following websites

The National Football League Savant Website

Football Statistics Downloadable Content.

The statistical methods that we used to analyze these data includes calculating the probability, hypothesis testing, linear regressions, etc. We were able to manipulate the databases into specific football situations and use R to perform the statistical analysis. The finding of this study was very promising.

Peer to Peer Social Networking

Travis Thurn

Mentor(s): Shuhui Yang

Undergraduate

Oral Presentation

Online social networking has become an integral part of daily living. Small businesses are encouraging customers to follow them on Facebook. Revolutions around the world are being live tweeted. Humanity has become more connected; however, there is an illusion of a personal connection but the reality is all communication goes through a middle man, the service provider. This centralized client-server model has a serious privacy flaw, no matter how private the content may be the provider reads and processes all the data in order to sell targeted advertising. There is also the problem of the network going down due to forces of nature, which leads to the failure of the social network system and hence the communication of people.

Our proposal is to rethink how online social networking occurs, from having a corporation handle all transactions to a direct connection between individuals and the network at large. Peer-to-peer (P2P) networking is a concept that has been in existence at least as early as 1969 and was popularized through file sharing in the late 1990s. It offers the privacy of direct communications between peers and the robustness to function in isolation. We have designed an amorphous model for the P2P infrastructure based social networks that the network changes as connections are made and dissolved. Based on this model, we propose a 2-phase communication protocol, which includes a swarm procedure and a distributed hashing procedure. Simulations are undergoing to verify the effectiveness and efficiency of the proposed design.

Directory: research -> wp-content -> uploads -> sites
uploads -> [Information Interfaces and Presentation
uploads -> Using Web Annotations for Asynchronous Collaboration Around Documents
uploads -> Supporting Email Workflow Gina Danielle Venolia, Laura Dabbish, jj cadiz, Anoop Gupta
uploads -> Efficient Image Manipulation via Run-time Compilation
uploads -> Vassal: Loadable Scheduler Support for Multi-Policy Scheduling
uploads -> Strider GhostBuster: Why It’s a bad Idea For Stealth Software To Hide Files
uploads -> High Performance Computing: Crays, Clusters, and Centers. What Next?
uploads -> Universal Plug and Play Machine Models
uploads -> An Abstract Communication Model
sites -> Faculty Proceedings April 3-4, 2018

Download 403.55 Kb.

Share with your friends: