*Last updated on November 20, 2007*

Course announcements (most recent first) are as follows:

- A timetable and evaluation sheet for the class presentations on Monday November 26 is available on the handouts page.
- A study guide is available on the handouts page. This consists of a 4-page document with question topics for the short-answer section of the final exam and space for notes.
- 5 practice multiple choice questions are also available on the handouts page.
- Office hour Monday November 12 is cancelled.
- The final Tayko solution is available as an Excel spreadsheet on the data page and also as a handout. This is for information only - the analysis from the second half of class on Wednesday October 31 went a little beyond what will be covered in homeworks and the final exam (although the concept of "oversampling" is important and could come up in the exam - this is covered in the textbook on p68-9).
- When working with large spreadsheets in Excel you can sometimes speed things up by replacing any formulas (e.g., for calculating dummy indicator variables) with "values" (i.e., copy the column containing formulas and "paste special > values" over the existing column of formulas). Then, every time Excel (or XLMiner) does something it won't automatically recalculate all the old formulas. Alternatively, you can change the default behavior for "calculating automatically" in the "options" (under "tools" I think) - then you have to press F9 to make Excel do calculations manually.
- There a few errors in chapter 5 of the textbook:
- p143: standard error formula should be
sqrt[(p
_{1}(1-p_{1})/N_{1})+(p_{2}(1-p_{2})/N_{2})] - p147: standard error formula should be
sqrt[(p(1-p)/N)+((p+d)(1-p-d)/N)] and final answer should be N =
0.096796/(0.00102)
^{2}= 92,963 - page 151: there should
*not*be a square root in the chi-square(x) formula

- p143: standard error formula should be
sqrt[(p
- There a few errors in chapter 6 of the textbook:
- p183: the first split children have variance
(0.9
^{2}+9*0.1^{2}) / 10 = 0.09. - p193: the second rule should result in a "milk" prediction, while the fourth rule should result in a "diet soda" prediction.

- p183: the first split children have variance
(0.9
- To complete the assignments you will need the full (unlocked) XLMiner software - see the syllabus or installation instructions here. If you get a message saying "#rows in the data set cannot exceed maximum allowed 600 rows" when trying the first assignment then you have not unlocked the software correctly.
- Additional resources are available at the textbook web-site.

*The views and opinions expressed in this page are strictly those
of the page author. The contents of this page have not been reviewed
or approved by the University of Oregon.*