ml intro

Question 1

The intersection of two convex sets is always convex.

Answer 1

TRUE

Question 2

The union of two convex sets is always convex.

Answer 2

FALSE

Question 3

The function f(x)=7+3x1+999x2 is linear.

Answer 3

FALSE

Question 4

A logistic function determines the threshold according to which the output of a linear model is classified.

Answer 4

FALSE Because YOU decide the threshold.

Question 5

In order to predict y, x2 is more informative than x1.

Answer 5

FALSE It’s exactly the other way around: X1 is more informative. This is because:
I(x1|y)=1/4log(1/4/(1/4+1/2))+…+…+…=1/4+1/4+1/4+1/4=1

Question 6

The mutual information between x1 and y is 1.

Answer 6

TRUE Because we have exactly 2 0 and 2 1.

Question 7

The linear correlation coefficient between xi and yi may change if x values are centered by subtracting their mean value μ in the following manner: xi−μ

Answer 7

FALSE Centering does not affect the linear correlation

Question 8

The entropy of a uniform probability distribution of n events is log2(n).

Answer 8

TRUE

Question 9

The use of Chi-square test to deny statistical independence means that, for example, a term in a phrase should not be used as a feature because the square of the number of nearest neighbors is too large.

Answer 9

FALSE No relationship between Chi-square test and nearest neighbors.

Question 10

If the Pearson correlation coefficient between two data features is zero, then such features are independent.

Answer 10

FALSE Pearson correlation ahd Mutual information are not related. They measures two completely different things.

Question 11

If the Pearson correlation coefficient between two data features is zero, the Mutual Information between such features is also zero.

Answer 11

FALSE

Question 12

If the number of input variables is 99, and one starts training from 55 different examples, the parameters of the linear model obtaining zero error on the examples can always be determined.

Answer 12

TRUE

The number of input variables need to be equal or greater than the number of examples for interpolation.

Question 13

The measurements of two phenomena are different in a statistically significant way if one can demonstrate in a theorem that the two measurements will never be equal.

Answer 13

FALSE

Because we only care of statistically significant results, not theorems.

Question 14

A result is statistically significant when it is obtained by democratic means, asking for the opinion of the largest possible number of experts.

Answer 14

FALSE

Question 15

For every k>0k>0, in leave-one-out cross-validation, one of the k partitions is left out as validation data and the other partitions are used as training data.

Answer 15

FALSE

Question 16

Goodness functions take measurements as input. No

Answer 16

TRUE

Question 17

Measurements can be only numerical.

Answer 17

FALSE

Question 18

Standard mathematical optimization requires the
existence of goodness functions to be optimized,
and the de nition of these functions is usually
feasible in the real world.

Answer 18

FALSE

Question 19

Machine learning requires the de nition of a

goodness function to be optimized

Answer 19

FALSE

Question 20

Machine learning techniques can build good

models only if abundant data is available.

Answer 20

TRUE

Question 21

The goodness function to be optimized is Gold(x),
the quantity of gold extracted from a mine at
position x.

Answer 21

TRUE

Question 22

The measurements are the gold quantities

extracted in di􀃗erent points.

Answer 22

FALSE

Question 23

Kriging is an example of descriptive analytics. No

Answer 23

FALSE

Question 24

The model of the goodness function, built using
measurements, aims at generalizing the results
obtained during the experiments.

Answer 24

TRUE

Question 25

The output of the model of an unknown point is
the average of the known values of its neighbors,
weighted by the neighbors’ distance to the
unknown point.

Answer 25

TRUE

Question 26

Images are an example of structured data.

Answer 26

FALSE

Question 27

Text is an example of unstructured data.

Answer 27

TRUE

Question 28

Vectors are an example of structured data.

Answer 28

TRUE

Question 29

Comma Separated Values (CSV) are an example of structured data.

Answer 29

TRUE

Question 30

Predictive analytics goes further than descriptive

and prescriptive analytics.

Answer 30

FALSE

Question 31

Descriptive analytics involves the analysis of

historical data in order to provide useful insights.

Answer 31

TRUE

Question 32

Predictive analytics tries to anticipate the e􀃗ects of
decisions, by creating models based on historical
data

Answer 32

TRUE

Question 33

Prescriptive analytics uses optimization and
simulation algorithms to consider the e􀃗ect of
many possible decisions, in order to take the
decision that it is expected to optimize a certain
goodness function.

Answer 33

TRUE