machine learning

Which statement is true about the K-Means algorithm?

All attributes must be numeric°

Since policies are only partially ordered, there can be many different optimal policies with different value functions.

False

You run gradient descent with learning rate ALPHA = 0.3 for it iterations and compute E(w) after each iteration. You find that E(w) increases over time. Which conclusion seems most plausible?

It would be better to decrease the learning rate

Regression differs from classification in that we...

Learn a function with numeric output

The K-means algorithm terminates when...

The clusters centers for the current iteration are identical to the cluster centers for the previous iteration.

Which of the following statements is not true about ordinal values?

The mean of such values can be calculated.

Which of the following is not true? Principal component analysis maps data into a lower dimensional space such that ...

The new features have a high correlation with the target output.

Which of the following statements about the bagging technique is not true?

The weak classifiers are weighted by the prediction accuracy

Which of these statements is true about nominal (categorical) values?

Their mode can be calculated in a meaningful way.

A support vector machine can deal with data that is not linearly separable by

Using a kernel function to learn a linear decision boundary in higher dimensional space than the input space and introduction a slack parameter for each example and minimising the total amount of slack.

A leaf node in a decision tree represents ...

a class

An internal node in a decision tree represents ...

a test specification

Supervised learning differs from unsupervised learning in that supervised learning requires ...

at least one output attribute

Which of the following cases indicates underfitting?

high training error, high test error

You are training a decision tree classifier with a minimal number of instances LaTeX: n n that are required for making a new decision node. The results from cross-validation indicate overfitting. To reduce overfitting you ...

increase n

Which of the following cases indicates overfitting?

low training error, high test error

Which of these is not a measure for the spread of the data?

mean

Which of these measures is most sensitive to outliers?

mean

In reinforcement learning we often need to balance exploitation vs. exploration. In this context, what is exploitation?

selecting the action with highest estimated value

Assume a classification problem with three classes A, B, and C. The false negatives of a classifier wrt. class A are ...

the instances in class A that are classified as B or C.

Assume a classification problem with three classes A, B, and C. The false positives of a classifier wrt. class A are ...

the instances in classes B or C that are classified as A.

What is the dimensionality of record data?

the number of attributes

Accuracy is a misleading measure for the performance of a classifier when ...

we do not have similar numbers of instances for each class