Decision Tree Tools Are Mainly Used In Research

A decision tree is a decision making tool, which uses graphic models to represent analysis of possible consequences in business decisions. This tool is mainly used in research, specifically in decision making. The possible consequences that the tool determines include, chance event outcomes, resource costs, and utility. In order to identify a strategy, the tool provides the feature of the use of decision, in more likely cases where a goal needs to be reached. Here we will discuss more decision tree examples.

The tree modeling is also used in cases where descriptive means for calculating conditional probabilities is required. A decision tree becomes a predictive model in case of data mining and machine learning. This is achieved by mapping from observations about an item to the conclusions reached about its target value. Decision tree models have got more descriptive names, such as, classification tree (discrete outcome), or regression tree (continuous outcome). The leaves, in the tree structure, represent the classifications, while the branches are the conjunctions of features leading to the classifications. The machine learning technique for inducing a decision tree from data is called decision tree learning, or just decision trees.

An influence diagram, which is a closely related model form of the decision tree in decision analysis, is often used as a visual and analytical decision support tool. In here the value of the competing alternatives are computed. As an example, let us take a factory manufacturing a product "B". The management has taken a decision to go into a new product line, and is required to take a decision between product "C" or "D". The company cannot take up both because of budget constraints. Analysis shows that, the product "B" would require an investment of $2m in research and design, with only 50% chance of the product to be successful. Even if the R&D brings out the product, the product will have 30% sales possibilities, with $5m in profit, and increasing the sales to 40% would bring in a profit if $10m, having 30% chance of no sales. Product "D" has a $2m cost in R&D, and has a 80% chance in sales with a profit of $10m, with 20% chance of no sales. The manufacturing cost of either of the products is $3m.

The question is which the better investment to achieve the maximum value is. If the situation could be taken up for analysis, decision tree would provide you with the decision making possibilities of the best strategy that you can take. The decision making tool would provide you with the alternatives, probabilities, payoffs, and resulting expected value calculations. Iin the above case products "C" and "D" would turn out a profit, with product "D" offering a higher return by $1m.

There are distinct advantages in using decision trees and influence diagrams, and these advantages include:

The tree is simple to understand and easy to interpret. The decision tree models could be well understood after a brief explanation of the process. The tree can produce meaningful value even with little hard data. Based on expert observation on the alternatives, probabilities, costs, and their preferences for outcomes, several important insights could be generated by decision trees. It facilitates the use of a white box model. The decision tree technique can be combined with other techniques. Decision tree models are used to optimise investment portfolios.

In the current use of data mining and machine learning, a decision tree is the most popular classification algorithms. There are tutorials which one can use as a self-contained introduction to the terminology of data mining, and you can learn without having to refer to many statistical and probabilistic pre-requisites. Decision tree application will surprise you with data mining applications, and if you are new to the application, you will find how easy it is.

Privacy Policy