| Line 2: | Line 2: | ||
=Details of Lecture 27, [[ECE662]] Spring 2010= | =Details of Lecture 27, [[ECE662]] Spring 2010= | ||
| − | April | + | April 23, 2010 (Make up class) |
In Lecture 27, we discussed [[Decision Trees|decision trees]]. We began with an example of decision tree where the first query does not decrease the impurity of the data (the country question as part of deciding if a person is married or not). We then gave another example of decision tree (the fruit identification tree) where each of the queries decreases the impurity of the data. That latter example featured several categories (the fruits), while the former featured only two categories ("married" and "not married"). We mentioned that decision trees should be build in such a way as to decrease the "impurity" of the data, and concluded by defining different measures of data impurity. | In Lecture 27, we discussed [[Decision Trees|decision trees]]. We began with an example of decision tree where the first query does not decrease the impurity of the data (the country question as part of deciding if a person is married or not). We then gave another example of decision tree (the fruit identification tree) where each of the queries decreases the impurity of the data. That latter example featured several categories (the fruits), while the former featured only two categories ("married" and "not married"). We mentioned that decision trees should be build in such a way as to decrease the "impurity" of the data, and concluded by defining different measures of data impurity. | ||
| − | Recall that next Thursday's lecture (4-29-10) is canceled, and that there is a make up class Friday (4-30-10), 1:30-2:30 in EE117. | + | Recall that next Thursday's lecture (4-29-10) is canceled, and that there is a make up class next Friday (4-30-10), 1:30-2:30 in EE117. |
Previous: [[Lecture26ECE662S10|Lecture 26]] | Previous: [[Lecture26ECE662S10|Lecture 26]] | ||
Latest revision as of 10:18, 27 April 2010
Details of Lecture 27, ECE662 Spring 2010
April 23, 2010 (Make up class)
In Lecture 27, we discussed decision trees. We began with an example of decision tree where the first query does not decrease the impurity of the data (the country question as part of deciding if a person is married or not). We then gave another example of decision tree (the fruit identification tree) where each of the queries decreases the impurity of the data. That latter example featured several categories (the fruits), while the former featured only two categories ("married" and "not married"). We mentioned that decision trees should be build in such a way as to decrease the "impurity" of the data, and concluded by defining different measures of data impurity.
Recall that next Thursday's lecture (4-29-10) is canceled, and that there is a make up class next Friday (4-30-10), 1:30-2:30 in EE117.
Previous: Lecture 26 Next: Lecture 28
