CS 1110 Final: CS 100M Cornell 2004sp Final Examst3solcs100jsp01

• School: Cornell University
• Department: Computer Science
• Course: CS 1110
• Professor: All

Document Summary

2. (3) when is the linearity assumption obviously false: when the x variable has 3 levels b. When the x variable has < 3 levels: when the y variable has 3 levels d. when the x variable has < 3 levels. 3. (3) which kind of model does occam"s razor tend to prefer: simple b. complex c. normal d. non-normal. 4. (3) which is the conditional mean function for the data generating process: 0 + 1x + b. 5. (3) what does the gauss-markov theorem say: maximum likelihood estimators are good, ordinary least squares estimators are good, linear estimators are good, unbiased estimators are good. 6. (3) suppose your n data values are produced by yi = 0 + 1xi + i, with all assumptions satisfied. Holding all else fixed, what reduces variance of: smaller n b. smaller 1 c. larger var(x) d. larger var( )