Blog

A nurse is caring for a client who has returned to the unit following a surgical procedure. The client’s oxygen saturation is 85%. Which of the following actions should the nurse take first?

When should you use a supervised learning method for your analysis?

This refers to the Camping affinity analysis script and output.A customer already has Snack_trailmix in his/her cart. (There may or may not be other items in there as well.) What additional item(s) would you recommend and why? Make sure you tell me which areas of the output you used to derive your answer.

Look at the figure below. It’s the plotcp from some random tree (not part of this exam.) Using just what you can tell from this data, which cp would be recommended for tree pruning? Enter your answer to as many decimal places as you can. cpsizeoftreeplot.png

In this course you’ve been exposed to two different supervised machine learning analysis methods that have “regression” in their title.  One is linear regression models (Lesson 3) and the other is Regression Trees, a subset of decision trees.How do they differ?  When can / should you use each one?

What is unsupervised learning?

This refers to the Camping affinity analysis script and output.What does your Rule 3 say? What sorts of business recommendations would you make if you were going to implement this rule?Make sure you tell me which areas of the output you used to derive your answer.

This question refers to the Classification Tree information presented above. Your boss doesn’t understand R code, and has asked you to give a general overview of what is going on here. Write a 1-2 paragraph overview in response to the question. Make sure you include the overall objective of this analysis, and any dead ends or good results.

You have a data set with the following variables: Household income, which runs from $30,000 to $120,000 Number of children in family, from 0 to 5 Number of years in current location, from 0 to 20+ Whether they rent or own, as a text field Number of miles driven per year, from 0 to 50,000 Population in their location, from 100 to 1 million + Where possible, you have standardized the variables, and you have recoded the rent/own text field into a binary 0 (for rent) and 1 (for own). You want to run a k-means algorithm on this entire data set, to try to determine different demographic niches. For example, you may want to separate out urban apartment-dwellers from rural retirees. Is it possible to run k-means clustering on all of these data fields? Say you have loaded the standardized variables above into columns 1 through 6 in a data frame called responses. Let’s say you want to try for 7 clusters.  In particular, can you do something like this? > fit

A healthy diet should include no more than 1% saturated fat and 7% trans fat. Both saturated and trans fats are EQUALLY a harmful to arteries.