Assume you want to perform supervised learning and to predict the price of houses according to the size of houses, it is an example of a clustering problem. 

(NOTE: Deducted points for the wrong answer(s). 0 point for 'Not Answering').

For the learning rate in gradient descent, if it is too big , it will require a lot of training time (assuming there is a convergence).

(NOTE: Deducted points for the wrong answer(s). 0 point for 'Not Answering').

Choose methods for evaluating features for feature selection.

(NOTE: Multiple answers are allowed).

From the code snippet for a GridSearchCV based cross-validation for Ridge alpha

hyper-parameter, choose all the correct

answers.

Like the name, the Logistic Regression algorithm mainly will be used to solve a regression problem.

(NOTE: Deducted points for the wrong answer(s). 0 point for 'Not Answering').

Among the k-Fold cross validation (CV)  and leave-p-out CV techniques, which one belongs to the exhaustive (or more exhaustive) method.

(NOTE: Deducted points for the wrong answer(s). 0 point for 'Not Answering').

In CV (Cross-Validation) process, which error is more important for the choice of the best parameters (including hyper-parameters)?

(NOTE: Deducted points for the wrong answers. 0 point for 'Not Answering').

The information of the weight table (weight  and variance

estimates) can be visualized in a weight plot.

The following plot shows the results from a

Choose all correct interpretation/explanations.

In CNN (Convolutional Neural Networks) model training phase, how the windows size and  the window stride in the Pooling layer can be built/selected?

From the code snippet for a GridSearchCV based cross-validation for multilayer perceptron (MLP) hyper-parameters, choose all the correct answers (assuming that the number of independent variables/features could be a large number (e.g., 100), and the number of instances are relatively small (e.g., 500)).