Projects (1999 - Spring 2008)

Engineering >> Computer Science & Engineering

by Brett Waugh

Submitted : Fall 2015

After collecting data about vulnerability ratings, the data was reworked into a desired format of [0,1] and in an ascending order of security. The adjusted data was used to create points on a plane then used in Least-Squares Regression. This left a large polynomial of three unknown variables. Taking the partial derivative with respect to each variable created three new equations to form a system of three equations and three unknown variables, which were placed back into an equation for a plane to give a model plane of the three data groups. With this new model equation the Pearson’s Correlation Coefficient was used with the known data and the data taken from the model. Once the Pearson’s Correlation Coefficient was known, squaring this term gave the coefficient of determination.

The Pearson’s Correlation Coefficient is r = 0.76146 which is a somewhat strong, positive correlation. The coefficient of determination is r^2=0.579821 which means that roughly 58% of the total variation can be described using the equation f(x,y)= -1.95204 x-0.148936 y+0.673732. Putting this into context, the strong correlation between the data suggests that the data is related and as one rating grades something higher, the others will, as well. The coefficient of determination says that the model can determine variability in the scores about 58% of the time.

[ Back ]

 Advisors : Arcadii Grinshpan, Mathematics and Statistics Jonathan Burns, Mathematics and Statistics Jarred Ligatti, Computer Science & Engineering Suggested By : Jarred Ligatti