Wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Lorenzo

Total questions: 26

Worksheet time: 16mins

Name
Class
Date
1.

An algorithm iteratively learns how to classify data by maximising the accuracy of its performance. However, what would likely happen when the dataset is highly unbalanced (e.g., 99% of positives and 1% of negatives)?

a)

the algorithm is not able to learn because the accuracy will always stay too low (~0.01)

b)

the algorithm is not able to learn because the accuracy will always stay too high (~0.99)

c)

accuracy is never related to the balancement of the classes

d)

I don't know

2.

An algorithm iteratively learns how to classify data by maximising the accuracy of its performance. However, what would likely happen when the dataset is highly unbalanced (e.g., 99% of positives and 1% of negatives)?

a)

the algorithm is not able to learn because the accuracy will always stay too low (~0.01)

b)

the algorithm is not able to learn because the accuracy will always stay too high (~0.99)

c)

accuracy is never related to the balancement of the classes

d)

I don't know

3.

You are developing a model for sick patient detection (sick=positive, healthy=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

4.

You are developing a model for sick patient detection (sick=positive, healthy=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

5.

You are developing a model for email spam detection (spam=positive, non-spam=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

6.

You are developing a model for email spam detection (spam=positive, non-spam=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

7.

You are developing a model for fraudolent transaction detection (fraudolent=positive, non-fraudolent=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

8.

You are developing a model for fraudolent transaction detection (fraudolent=positive, non-fraudolent=negative). Which metric would you use?

a)

Accuracy

b)

Precision

c)

Recall

d)

I don't know

9.

Which model would you choose?

a)

A

b)

B

c)

I don't know

10.

Which model would you choose?

a)

A

b)

B

c)

I don't know

11.

You are developing a model that fits the luminosity profile of a Galaxy. You know that your image can be affected by cosmic rays, hot pixels, dead pixels. Which metric would you use?

a)

MSE

b)

MAE

c)

F1-score

d)

I don't know

12.

You are developing a model that fits the luminosity profile of a Galaxy. You know that your image can be affected by cosmic rays, hot pixels, dead pixels. Which metric would you use?

a)

MSE

b)

MAE

c)

F1-score

d)

I don't know

13.

When is cross-validation preferable over train/test split?

a)

On small datasets

b)

On large datasets

c)

Cross-validation is always preferable

d)

I don't know

14.

When is cross-validation preferable over train/test split?

a)

On small datasets

b)

On large datasets

c)

Cross-validation is always preferable

d)

I don't know

15.

Which one of the following is the best strategy (typically)?

a)

Very small eta

b)

Very high eta

c)

Eta starts small and gets bigger and bigger at each iteration

d)

Eta starts big and gets smaller and smaller at each iteration

e)

I don't know

16.

Which one of the following is the best strategy (typically)?

a)

Very small eta

b)

Very high eta

c)

Eta starts small and gets bigger and bigger at each iteration

d)

Eta starts large and gets smaller and smaller at each iteration

e)

I don't know

17.

Which is the main pitfall of Stochastic descent?

a)

It is too fast to converge to the minimum

b)

It is too slow to converge to the minimum

c)

It never settles at the minimum

d)

I don't know

18.

Which is the main pitfall of Stochastic descent?

a)

It is too fast to converge to the minimum

b)

It is too slow to converge to the minimum

c)

It never settles at the minimum

d)

I don't know

19.

Do all optimizers lead to the same model if you let them run long enough?

a)

Yes

b)

No

c)

I don't know

20.

Do all optimizers lead to the same model if you let them run long enough?

a)

Yes

b)

No

c)

I don't know

21.

How do you fix that?

a)

I use the Batch Gradient Descend

b)

I don't know

c)

I reduce the batch size

d)

I increase the batch size

e)

I decrease the learning rate

22.

How do you fix that?

a)

I use the Batch Gradient Descend

b)

I don't know

c)

I reduce the batch size

d)

I increase the batch size

e)

I decrease the learning rate

23.

Which is the model that is overfitting?

a)

Left panel

b)

Central panel

c)

Right panel

d)

I don't know

24.

Which is the model that is overfitting?

a)

Left panel

b)

Central panel

c)

Right panel

d)

I don't know

25.

Suppose you are using Ridge Regression and you notice that the training error and the validation error are almost equal and fairly high.

a)

It is ok as long as the two errors are similar

b)

I increase the regularization hyperparameter

c)

I decrease the regularization hyperparameter

d)

I don't know

26.

Suppose you are using Ridge Regression and you notice that the training error and the validation error are almost equal and fairly high.

a)

It is ok as long as the two errors are similar

b)

I increase the regularization hyperparameter

c)

I decrease the regularization hyperparameter

d)

I don't know