ISYE-ISYE-6501 Week 1 Assignment

Document Content and Description Below

ISYE-ISYE-6501 Week 1 Assignment Question 2.1 Describe a situation or problem from your job, everyday life, current events, etc., for which a classification model would be appropriate. List some pre ... dictors that you use. Designing user personas in product design developments The goal of user personas is to develop realistic representations of key audiences that give a clear picture of their expectations and use cases for products. Using user personas in the product design process is to understand user characteristics, needs, goals, etc. and gain valuable insights into user journeys, and later, test prototypes. With user personas, product owners, designers and engineers should be able to do the following: ● Put users at the center of product design to build empathy ● Focus on their campaign process and requirements ● Share insights with other stakeholders to gain consensuses ● Ultimately make products more actionable and desirable aligning users' needs and goals K-means classification would be an appropriate model to identify how many user persona types should be identified by using the current product analytics data that illustrate product uses. Potential predictors for this model may include 1) number of product access, 2) frequency of using specific product features (these features should signify what task each user should accomplish using the product), 3) 2 data entries for agency and department identifications, 4) exporting specific data Question 2.2.1-2 Using the support vector machine function ksvm contained in the R package kernlab, find a good classifier for this data. Show the equation of your classifier, and how well it classifies the data points in the full data set. ● Executive data finding summary ● Best accuracy - 98.16514% with rbfdot & C=1,000 C = 0.01 C = 1 C = 10 C = 100 C = 1,000 polydot 0.8639144 0.8639144 0.8639144 0.8639144 0.8623853 rbfdot 0.8593272 0.8715596 0.9082569 0.9571865 0.9816514 vanilliabot 0.8639144 0.8639144 0.8639144 0.8639144 0.8623853 > data=read.table("~/Desktop/GATech/ISYE-6501/Week1/credit_card_data-headers.txt",header = T,sep='\t') > library(kernlab) > model1v <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="vanilladot", C=0.1, scaled=TRUE) Setting default kernel parameters > model2v <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="vanilladot", C=1, scaled=TRUE) Setting default kernel parameters > model3v <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="vanilladot", C=10, scaled=TRUE) Setting default kernel parameters > model4v <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="vanilladot", C=100, scaled=TRUE) Setting default kernel parameters > model5v <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="vanilladot", C=1000, scaled=TRUE) Setting default kernel parameters > model1p <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="polydot", C=0.1, scaled=TRUE) Setting default kernel parameters > model2p <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="polydot", C=1, scaled=TRUE) Setting default kernel parameters > model3p <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="polydot", C=10, scaled=TRUE) Setting default kernel parameters > model4p <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="polydot", C=100, scaled=TRUE) Setting default kernel parameters > model5p <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="polydot", C=1000, scaled=TRUE) Setting default kernel parameters > model1r <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="rbfdot", C=0.1, scaled=TRUE) > model1r <- ksvm(as.matrix(data[,1:10]), as.factor(data[,11]), type="C-svc", kernel="rbfdot", C=0.1, scaled=TRUE) [Show More]

Last updated: 3 years ago

Preview 1 out of 9 pages