Crowdly

minimizes total error

minimizes variance

What is the Turing test?

a test of machine’s ability to exhibit intelligent behavior similar to human’s

a test of machine’s ability to express emotions

a test of machine’s ability to grow self-consciousness

a test in which a human evaluator is judging a conversation between human and a machine

The F1 score:

is a harmonic mean of accuracy and precision

is a harmonic mean of accuracy and recall

is used in evaluation of classification models

is a harmonic mean of precision and recall

We want a model to determine which items in a supermarket are most commonly bought together based on purchase data from cash registers. This is:

an association rule learning problem

an unsupervised learning problem

a classification problem

a supervised learning problem

What does cross-coupling in a perceptron machine mean?

connections between units within the same layer of a perceptron machine

connections between non-consecutive layers of a perceptron machine

connections from units in a later layer to units in a previous layer of a perceptron machine

Which of the following statements about unsupervised learning are true?

Data is unstructured and no labelling is needed.

Datasets are usually cheaper to collect, however they are much bigger than SL training sets.

We don’t need separate test and validation sets.

What is a simple linear regression?

a linear regression with two independent variables

a linear regression with two quantitative variables, one independent and one dependent

a linear regression with only one quantitative variable

a linear regression with two dependent variables

What is plurality voting in K-Nearest Neighbors algorithm?

a method of assigning a new sample a class label that is most frequently represented among it’s k nearest neighbors

50%

a method of assigning a new sample a class label that is the average of labels of it’s k nearest neighbors

50%

a method of assigning a new sample a class label that is represented by more than 50% of it’s k nearest neighbors

The ratio is called

recall

accuracy

precision

specificity

Which of the following problems are typical for supervised learning?

association

clustering

classification

linear regression

anomaly detection