ECON 2000 UU Variety of Models to Predict if A Person Got Vaxx Program

User Generated

wnzobb

Computer Science

ECON 2000

umbrella university

ECON

Description

Unformatted Attachment Preview

Skip to content request kevinrfoster/ecob2000_lab7 Public • • Watch 1 Star0 • Fork 2 • • • • • • • • • Code Issues Pull requests Actions Projects Wiki Security Insights main 1 branch 0 tags Go to file Add fileCode Latest commit kevinrfoster Add files via upload 2fb2bdd6 days ago Git stats • 6 commits Files Latest commit message Add files via upload Add files via upload Add files via upload README.md Lab 7 Econ B2000, MA Econometrics Fall 2021 For this lab, we will estimate a variety of models to try to predict if a person got vaxx (same data as last week). Compare logit with OLS in terms of prediction and set up the variables to be ready to expand into other models (next week). And give me some better output, it’s time to stop dumping all your output into one file but instead get thoughtful about presenting results. First decide on how you’re defining your subgroup (all adults or 12+? Within certain age? Other?) then find some basic statistics – what fraction are not vaxxed? (Later go back to look at simple stats for subgroups to see if there are sharp differences.) Explain what you’re doing with NA. You did this last week (along with defining vaxx) so check back. You might do the same or choose to improve. Run several different types of models to explain vaccination rates with some explanatory variables, vaxx ~ TBIRTH_YEAR + EEDUC + MS + RRACE + RHISPANIC + GENID_DESCRIBE + REGION. Compare the confusion matrix for linear model and logit. Look at subgroups to see if there are particular groups where the models are more confused. Look at the tradeoff of false positive vs false negative. Are there explanatory variables (features) that are consistently of little predictive value? Can you find better ones? Are these X-variables exogenous? As you add more, think about causality. We want to set up the data in a way that is common to all of the models. Some of the estimation procedures are not as tolerant about factors so we need to set those as dummies. Some are also intolerant of NA values. I’ll show the code for the basic set of explanatory variables, which you can modify as you see fit. The R command model.matrix() creates a set of dummy variables out of a factor. Run this to see a for-instance: d_educ
Purchase answer to see full attachment
User generated content is uploaded by users for the purposes of learning and should be used following Studypool's honor code & terms of service.

Explanation & Answer

View attached ex...

Related Tags