- Build a classification model for a simple credit history evaluator based on these 14 training examples. In particular, assume we wish to assign a credit Risk (the Label) of high, moderate, low to people based on the following properties of their credit rating: Security: represents collateral with possible values {adequate, none) Salary: represents income with possible values {"0$ to $30K", "$30K to $60K", "over $60K"} Debt: with possible values {high, low} Reputation: represents credit history, with possible values {good, bad, unknown} Using Ross Quinlan's ID3 algorithm (the basis of C4.5 and J48), manually (by hand) build a Decision Tree from the given dataset. Show all of your work, calculations, and decisions as you build the tree. Training Examples obs Security Salary $0 to $30K Reputation bad 1 none 2 none $30 to $60K Debt high high low low unknown unknown Risk high high moderate high low 3 none $30 to $60K 4 none $0 to $30K unknown 5 none over $60K low low 6 adequate unknown unknown bad 7 none low low high moderate 8 over $60K $0 to $30K over $60K over $60K over $60K low adequate none bad 9 low low 10 adequate low 11 none $0 to $30K good good good good good had high high high high high high moderate 12 none $30 to $60K 13 low none over $60K 20 to 50V high decision nodes root node salary at least $50,000 yes no commute more than 1 hour yes decline offer no yes offers free coffee decline offer no leaf nodes Decision Tree: Should I accept a new job offer? accept offer decline offer - Build a classification model for a simple credit history evaluator based on these 14 training examples. In particular, assume we wish to assign a credit Risk (the Label) of high, moderate, low to people based on the following properties of their credit rating: Security: represents collateral with possible values {adequate, none) Salary: represents income with possible values {"0$ to $30K", "$30K to $60K", "over $60K"} Debt: with possible values {high, low} Reputation: represents credit history, with possible values {good, bad, unknown} Using Ross Quinlan's ID3 algorithm (the basis of C4.5 and J48), manually (by hand) build a Decision Tree from the given dataset. Show all of your work, calculations, and decisions as you build the tree. Training Examples obs Security Salary $0 to $30K Reputation bad 1 none 2 none $30 to $60K Debt high high low low unknown unknown Risk high high moderate high low 3 none $30 to $60K 4 none $0 to $30K unknown 5 none over $60K low low 6 adequate unknown unknown bad 7 none low low high moderate 8 over $60K $0 to $30K over $60K over $60K over $60K low adequate none bad 9 low low 10 adequate low 11 none $0 to $30K good good good good good had high high high high high high moderate 12 none $30 to $60K 13 low none over $60K 20 to 50V high decision nodes root node salary at least $50,000 yes no commute more than 1 hour yes decline offer no yes offers free coffee decline offer no leaf nodes Decision Tree: Should I accept a new job offer? accept offer decline offer
