Before class question: June 18
Spend 5 minutes to think about the following questions:
1. How can decision trees be used in businesses? What kind of business operations?
2. What do you think will be the difference between decision tree analysis and clustering analysis?
Yili Hong
Kevin Yili Hong
1. Decision trees can be very useful in business, if used correctly. For instance, a decision tree could be used by customer service representatives in dealing with a customer who returns a product to the store.
This would ensure that all customers are served uniformly and, because of the uniformity, certain ways of dealing with the business’s customers may be enhanced or eliminated easily.
This is just one example of a decision tree’s business application. Others may include a bank’s loan approval process, suggesting additional products to a customer, simplifying sales leads and many others, as well.
2. I think that a decision tree will be a tool that helps guide and suggest what actual choices should be made, whereas a clustering analysis seems to be a tool for analyzing data to notice patterns and does not suggest a course of action.
Can be used to make decision based on pre defined parameter. I
I think decision tree analysis is more conditional depending on the paths you make, on the other side I think clustering is a more detailed information about a certain group
I think that decision trees can be used as a protocol. Something that people will follow to make better decision (or try at least).
I think of decision trees as a script to a routine, while clustering analysis I believe will be based more on summarizing data to get information, I don’t really know..
Decision trees help analyze the outcomes of a few alternative actions before making a decision can help you determine if you’ve made a decision that produces the most favorable or least painful consequences. The decision tree is more helpful in selecting a choice that should be made, and clustering is more for analyzing data and is more detailed.
Decision trees would be useful for routine tasks such as troubleshooting because they follow a predefined course. They can make a suggestion but it will be based on percentages and not absolute so human judgment is needed. I also think an important point is that decision trees are only as useful as the person(s) that set it up and the logic employed. Clustering seems to be a tool that groups data based on similar attributes, this aggregated data can allow generalizations to be drawn
1. Financial analysts can use decision trees to determine the probability of bear or bull in response to say, Microsoft’s recent unveiling. The fact that there were no leaks makes the probability of bull higher; leaks about an unimaginative product will increase the probability of bear.
2. Decision trees graph possible outcomes while clustering is identifying similar items and grouping them together.
1. Decision trees can be used in marketing and business development operations. Businesses are constantly making decisions regarding product expansion, marketing, research and development, hiring new employees, etc., and if a decision tree is used all of these decisions can be viewed simultaneously. This allows a company to save a lot of time and effort, and see how things work in relation to each other.
2.Decision trees are a decision support tool, which graphs its decisions in a tree-like model or graph, while clustering is the task of assigning a set of objects into groups, so the objects on that cluster are all similar. The main difference between the two is that decision trees are broad and include all different types of data, while clusters are very specific and narrow.
Decision trees help analyzing the outcomes of few alternatives actions before making a decision. Decision trees are useful and can guide toward a most favorable decision.
In Business, decision trees can assist executives in making strategic decision. Very often, managers have to deal with certain situations and decisions where the consequences of their choices are unclear. They have trouble making decision and usually will use their intuition; which is not always the best.
A manager could benefit of a decision tree when making difficult decision:
-The first thing to do would be determining the decision that needs to be made for a specific situation.
-Second, collect all possible solutions to solve the problem.
-Then all possible consequences that could occur in event of any of the solution implemented.
-The final goal is to obtain probabilities of the likelihood of each consequence occurring.
These four steps follow the process of a decision tree.
2. Clustering is a data analysis task. Using statistical method, clustering analyze the data that identifies groups of sample data to act similarly. Clustering is very useful to use in Marketing when doing segmentation between different markets for instance. Data are then groups in different segment and study of the relationship between them can be made easily. Clustering groups data that behave similarly, study relationship between them and collect information out of it.
1. Decision trees are used in business such as expansion of a business, selection of location, and types of products you want to produce. It is a process that breaks down multiple decisions and place road blocks to disapprove or approve something.
2. Clustering is used to determine distinct groups of data, and it goes off of data across multiple dimensions. Decision Tree is used to classify data according to a pre-defined outcome and it goes off of characteristics of data.
A decision tree can be used to assist executives in making strategic decisions. It provides a visual representation of the choices facing an executive. It gives management the ability to evaluate all current information, weighing in the consequences of each “choice”, and allows them to develop a strategy with little interruption to their day-to-day business.
1. Decision trees can be used in businesses to assist in daily transactions such as flagging a credit card transaction as a fraudulent charge.
2. Decision trees seems to be more of a tool to guide the decision making process with suggestions, etc. where as cluster analysis is a tool that analyzes data and is more focused on classifying specific data
1. A decision tree can be used in most aspects of a business. They can be used in manufacturing (deciding which method of production is cost effective, marketing (which type of media to use), strategic planning/supply chain management (deciding what type of plant to build), finance, etc. A decision tree is a “map” that provides several “routes” (paths) that lead to all the possible destinations (decisions). It is in aid in making an informed decision.
2. Clustering is used primarily in marketing, especially when identifying customer segments. Cluster analysis allows marketers to better understand what “groups” customers fall in to or what preferences they like. These clusters then allow marketers to adjust/tailor media campaigns to more effectively target customers to receive a better response.
1) Decision Trees remind me of ERDs in the sense that they’re like blueprints; Decision trees help a firm, or group to create real life situations with possible solutions with contingency plans. And just like the ERD, if a problem arises later, the firm/group can go back and change the blueprint instead of having to change the whole system.
2) Clustering Analysis sounds like what it is, clustering data. Similar to grouping data using pivot tables, clustering helps with analysis of data and being able to identify what exactly the information being stored, means. The better the analyst can cluster, the more meaning the information will have, hopefully.