Jaclyn Hansberry

A decision tree could be used to help businesses find out which products in their company are the most profitable among consumers . The decision tree can answer questions on whether age plays a factor, or socioeconomic status. Will a product be more popular among one race or another? Cultures and Values? These questions can determine product profitability as well as how the company should target, place, and market certain products. The most important key component is choosing predictors that will influence the most positive (least error) outcome.
Reply

A decision tree can be used to get better estimates on what type of people are more likely to purchase store brand types of comestibles. Some good variables to use in this decision tree could be the income and whether or not the individual has children.

The questions would be should we write off an account payable or invoice. The decision tree could break down the factors or predictors that would help the business in knowing whether they should still expect payment or write it off and move on. To select the predictors you would want to use information that directly relates to repayment like how long has the account been past due, have they not paid in the past, how large is the balance, etc.

1. Where should a hospital devote resources to prevent Hospital Acquired Infections (HAIs)? The data I would collect are patient ID, diagnosis, procedure, equipment used, HAI acquired (Y/N), Type of HAI, if acquired.
2. I would advise to think first of the most direct variables related to the business problem at hand. If HAI’s are your business problem, you could choose the patient’s age, however HAI’s affect patients across all ages-therefore age is likely not a direct cause of HAIs. Instead focus on the process inputs like the procedure and equipment which are more likely to have a direct impact on the patient’s likelihood of contracting an HAI.

One business question we could answer is whether or not a person would eat at a new vegan restaurant. We could use factors such as gender, age, income, martial status, lifestyle, kids, etc. Some advice on choosing the right predictors is to chose the ones that are statistically significant. Basically, chose the relevant predictors that make a difference in the output. If a factor, such a person’s first name, has no effect on whether he or she will eat at the restaurant, then it shouldn’t be included in the decision tree.

Use decision tree to decide whether to deliver gym advertisements to people’s e-mail. We may use the data about their salary, age, height, weight and their work time, etc. In my opinion, salary is the first factor to be considered to be the predictors.
When we chose predictors, we should think about what factors can divide different groups significantly and properly.

(1) A possible business question that can be answered through a decision tree could be whether a company wants to expand by opening up a new business location. The predictor variables that I think would be needed are: how is the economy looking for the location you are interested in (that city perhaps), population in the location in which your business is interested, the demand for your product in that location, and what business competition would look like in that area.

(2) I think these predictors would be a good start for the decision tree but it would ultimately depend on more specific parameters depending on the business and where they aspire to reach as far as their goals are. I think the advice necessary for this situation would be to narrow down the specifics for that individual business and their market environment.

I would think a good example of using decision trees would be helping a school district decide if there needed to be more public school added to a particular district.
Business questions: Do more public schools (K-12) need to be added to the school district of Philadelphia County?

Possible predictors/variables wold be: Census data: household income, household members, ages of each; proximity to a public school for the household; proximity to private schools for the household; availability of bus routes (public/private) among others.

I would advise the collector of data to avoid obtaining information irrelevant to the analysis. Things such as SSN, Name and other personal information are irrelevant to this analysis. Also to canvas the designated area as completely as possible, using the most up to date census information. Outdated data will skew the results as population keeps growing and census screen are done only every 10 or so years.

A decision tree could be used when asked: Should I send my advertisement via email/
The data I would use is age, income, phone carrier with details ( could be null).
I would use age because many elderly would rather receive an advertisement through the mail. Income is important to decide whether they have the capability to check their email daily. Phone carrier details are important to see whether they can check their email regularly on the go.

To pick the right predictors you should split them until you get one statistically significant outcome.

A decision tree could be used to find which city someone should move to. The data needed for this could be job type, income, level of education, age, married, and number of children. To select the best predictor you should keep splitting until the data is together and the data is not so different from one another.

http://blog.revolutionanalytics.com/2014/05/companies-using-r-in-2014.html
Companies such as Google, Facebook, Linkedin, etc use R studio fr purposes of advertising and economic forecasting, big data visualization, and statistical analysis respectively and actually in their own right because as a scripting language, R studio proves to be a useful tool analyzing data

AirBnB is a top tier company that uses R packages to scale data science. In small data science teams, individual contributors often write single functions, scripts, or templates to optimize their workflows. As the team grows within the company, different people develop their own tools to solve similar problems. The package is developed in an internal Github Enterprise repository. There, users can submit issues and suggest enhancements. As new code is submitted in a branch, it is peer reviewed by an Rbnb developers group. Once the changes are approved and documented, they are merged into the master codebase as a new version of the package.

https://medium.com/airbnb-engineering/using-r-packages-and-education-to-scale-data-science-at-airbnb-906faa58e12d

So… In class, as we all know, there was lots of moaning and groaning about the ease of use of R. Well good news (maybe?) is that R, specifically Revolution Analytics, the R-language and data crunching specialist, has been acquired by Microsoft in 2015! The main task of Microsoft is to make the data mining giant more user friendly and eventually monetize it. While the aptly named, Microsoft R (Microsoft Data Science Virtual Machine), embraces the open source of the traditional, they are after all for profit company. R is currently being used by magnates such as Facebook, Twitter, Uber, AirBnB, IBM and Google for data analytics and there is lots of interest in the tech world to see what Microsoft has in mind for this new venture.

https://www.forbes.com/sites/adrianbridgwater/2016/01/15/microsoft-r-one-big-data-tool-to-rule-them-all/#94f3c5d651f4

https://www.quora.com/Which-organizations-use-R

The New York Times uses R for interactive features, data journalism, and data visualization. For example, they created a visualization to show Mariano Rivera’s baseball performance in comparison to other MLB pitchers’ performances. The graphic started as a hand-drawn sketch but then was created into a line-chart using R.

The article I read talked about R being used to help people in the financial services sector. R is adaptable and able to change with the needs of the user so it is perfect to use for financial services. It is also able to easily calculate everything needed for financial services and makes everything fast. The article I read links to a 40 minute talk about how this is helping to change financial services today.
https://www.r-bloggers.com/r-and-finance/