Chapter 6 Decision Trees PDF

Decision Trees Chapter 6 Learning Objectives Understand decision trees Identify key elements for constructing a decision tree Create a decision tree with simple dataset Identify popular decision trees algorithms Textbook: Data Analyics Made Accessible What...

Decision Trees Chapter 6 Learning Objectives Understand decision trees Identify key elements for constructing a decision tree Create a decision tree with simple dataset Identify popular decision trees algorithms Textbook: Data Analyics Made Accessible What are Decision Trees? One of the most widely used supervised classification technique Simplest way to guide one’s path to a decision Decisions can be simple or complex These are hierarchically branched structures Decision is arrived by asking question in the hierarchical order A good decision tree is short and arrives at a decision by asking the most relevant questions Can be generated from small datasets and applied to broader population Most convenient for simple binary decisions Textbook: Data Analyics Made Accessible Case Study: Predicting Heart Attacks using Decision Trees Heart attacks prediction using data mining The case study data is based on patients who have already suffered a heart attack The objective of the case was to predict which of the patients were at risk of dying from a second heart attack within the next 30 days Prediction would ultimately determine the treatment plan CART Algorithm was used Based on a set of input features (more than 100 variables) the output is predicted Data transformation and data cleansing performed before running the algorithm Blood pressure, age, and sinus problems were considered in asking questions The decision tree predicted 86.5% of the cases correctly. Textbook: Data Analyics Made Accessible Results The decision tree showed that if Blood Pressure was low (