Hey, I have some question regarding the following question:
1) is it linear regression problem or is it another problem? (linear classifier for example)
2) why C is the correct answer? we learned in another lecture about the problem of "under-determined task" when N << D, isn't it also here?
3) what is the meaning of "work" here exactly? is it enough on the training dataset or it should be a good generalizer for new inputs?
From the output values, it seems that this is a classification task, but the question decides to solve it with a regression model (with continuous outputs).
You are right that 'work' is rather vague here. I can only deduce what is meant by looking at the answers :) If we assume 'work' means 'fit the training data', being under-determined is not a problem.
Don't worry too much about this question. We'll try to make sure questions are well-defined.
2016
Hey, I have some question regarding the following question:
1) is it linear regression problem or is it another problem? (linear classifier for example)
2) why C is the correct answer? we learned in another lecture about the problem of "under-determined task" when N << D, isn't it also here?
3) what is the meaning of "work" here exactly? is it enough on the training dataset or it should be a good generalizer for new inputs?
Thanks in advance,
1
From the output values, it seems that this is a classification task, but the question decides to solve it with a regression model (with continuous outputs).
You are right that 'work' is rather vague here. I can only deduce what is meant by looking at the answers :) If we assume 'work' means 'fit the training data', being under-determined is not a problem.
Don't worry too much about this question. We'll try to make sure questions are well-defined.
Add comment