4 Responses

1. 

   Lorans · January 23, 2017 at 21:34:09 · →

   Please, go through the power-point again. I like it when the class goes through the power-point in the lecture, it make things easier to remember the more we go through things.

2. 

   K · January 23, 2017 at 21:34:53 · →

   Please explain the shannon plot, the sequence logo plot, and XOR problem again

   For overfitting, when to stop again? (The graph with testing dataset and training dataset ) Please explain again.

   Please explain sliding window techniques again.

   For TargetP, when you remove similar sequence, do u consider the entire sequence, even the pre-sequence part? Please explain the output table again please

3. 

   Ioanna · January 24, 2017 at 00:05:15 · →

   1.In Supervised Learning, what is the output of the given exmple? A yes or no answer? And what about the accuracy? Can the data always be used and the confidence interval just changes or is the confidence interval already set and the data can lead to prediction only if can be in the desired confidence interval?

   2.Could we have an example for Unsupervised Learning?

   3.In Support Vector Machines, how is the kernel type chosen? Is there an interpretation, e.g. biological etc behind this?

   4. Could we have some more details on the consequences of overtraining? Also an example would help.

   5.In the presequences in the example of subcellular localization could you please provide some more details regarding the weak consensus?

4. 

   V · January 24, 2017 at 09:10:31 · →

   Could we go through the Target P example again?
   I am not sure about the outcome and the reliability scores,
   Thanks