wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

ML

Total questions: 30

Worksheet time: 16mins

Name
Class
Date
1.

A computer program is said to learn from experience E with


respect to some task T and some performance measure P if its


performance on T, as measured by P, improves with experience E.


Suppose we feed a learning algorithm a lot of historical weather


data, and have it learn to predict weather. In this setting, what is the task T?

a)

The weather prediction

b)

The process of the algorithm examining a large amount of historical weather data

c)

The probability of it correctly predicting future weather

d)

None of the above

2.

What does a classification algorithm do?

a)

Predicts real number responses such as changes in temperature, date, or time

b)

Assigns data to a predefined category

c)

Clusters responses in groups based on similarity, to find patterns

d)

Compares predicted data classifications to the actual class labels in the data

3.

Suppose you are working on weather prediction, and your weather station makes one of three predictions for each day's weather:

Sunny, Cloudy or Rainy. You'd like to use a learning algorithm to predict tomorrow's weather.

Would you treat this as a classification or a regression problem?

a)

Regression

b)

Classification

4.

In a typical machine learning problem, what's the right order for the following decisions:

a)

1- Choose a hypothesis function

2- Choose a cost function

3- Choose an optimization algorithm

b)

1- Choose an optimization algorithm

2- Choose a cost function

3- Choose a hypothesis function

c)

1- Choose a cost function

2- Choose an optimization algorithm

3- Choose a hypothesis function

d)

1- Choose a cost function

2- Choose a hypothesis function

3- Choose an optimization algorithm

5.

The gradient descent is:

a)

A cost function

b)

An hypothesis function

c)

An optimization algorithm

6.

What is overfitting?

a)

When a predictive model is accurate but takes too long to run

b)

When you apply a powerful deep learning algorithm to a simple machine learning problem

c)

When the model learns specifics of the training data that can’t be generalized to a larger data set

d)

When you perform hyperparameter tuning and performance degrades

7.

What kind of table compares classifications predicted by the model with the actual class labels?

a)

Chaos table

b)

Prediction plot

c)

Confusion matrix

d)

Residual plot

8.

The goal of the gradient descent is:

a)

To find the set of parameters that minimizes the cost function

b)

To find the set of parameters that maximizes the cost function

9.

The Learning Rate is an hyper parameter in Gradient Descent used to update the parameters in the optimisation problem, which of the following is true?

a)

If it's too small, the convergence will be slower

b)

If it's too small, the convergence will be quicker

c)

If it's too large, we are sure to find the local optimum

d)

If it's too large, we may overshoot the local optimum

10.

In logistic regression, the output of the model is:

a)

A probability between 0 and 1

b)

Either 0 or 1

c)

Any real number

11.

Which of the following is true:

a)

In linear regression, we use the Least Square function as a cost function.

b)

In linear regression, we use the Log Loss function as a cost function.

c)

In logistic regression, we use the Least Square function as a cost function.

d)

In logistic regression, we use the Log Loss function as a cost function.

12.

The log loss function is defined as:

a)
b)
13.

Decision Trees Algorithm can be used for

a)

Regression only

b)

Classification only

c)

Both

14.

The technique in which we grow a tree to its maximum size, then remove some of its sub-nodes is called

a)

Un-Splitting

b)

De-Splitting

c)

Pruning

15.

The R2 score is better suited for:

a)

Regression tasks

b)

Classification tasks

16.

In classification, a False Positive is when:

a)

The model predicts TRUE and the real observation is TRUE

b)

The model predicts TRUE and the real observation is FALSE

c)

The model predicts FALSE and the real observation is TRUE

d)

The model predicts FALSE and the real observation is FALSE

17.

If we are building a cancer detection model (predict TRUE if cancer, FALSE if not), we would rather want to minimize:

a)

The false positive errors (Type 1 errors)

b)

The false negative errors (Type 2 errors)

18.

If we are building a spam detection model (predict TRUE if spam, FALSE if not), we would rather want to minimize:

a)

The false positive errors (Type 1 errors)

b)

The false negative errors (Type 2 errors)

19.

If we are dealing with a highly unbalanced dataset (0.01% of TRUE labels) we should prefer:

a)

The accuracy (Number of correct predictions / total)

b)

The recall (Number of correctly predicted positives / total of actual positives)

20.

Question 2
Suppose you ran regularized logistic regression twice, once with  λ=0\lambda=0  , and once with  λ=1\lambda=1   . You get two different   θ\theta parameters:  θ1=[74.81 , 45.05]\theta_1=\left[74.81\ ,\ 45.05\right]  and 

 θ2=[1.37 , 0.51] \theta_2=\left[1.37\ ,\ 0.51\right]\  

Which one corresponds to  λ=1\lambda=1  ?

a)

 θ1=[74.81 , 45.05]\theta_1=\left[74.81\ ,\ 45.05\right]  

b)

 θ2=[1.37 , 0.51] \theta_2=\left[1.37\ ,\ 0.51\right]\   

21.


In which one of the following figures do you think the hypothesis has overfit the training set?

a)
b)
c)
d)
22.

In which one of the following figures do you think the hypothesis has underfit the training set?

a)
b)
c)
d)
23.

Which one is true ?

a)

The more complex a model is, the lower its Bias and the higher its Variance.

b)

The more complex a model is, the higher its Bias and the lower its Variance.

24.

Which of the following diagrams represents the high bias, low variance situation?

a)
b)
c)
d)
25.

Which of the following diagrams represents the low bias, high variance situation?

a)
b)
c)
d)
26.

Regularization helps:

a)

To decrease the model's bias

b)

To decrease the model's variance

c)

To decrease the model's complexity

27.

What's the difference between the model's parameters and the model's hyper-parameteres?

a)

The model's parameters are learned during the training and the model's hyperparameters are specified outside of the training

b)

The model's parameters are learned outside of the training and the model's hyperparameters are specified during the training

28.

Which of the following feature engineering techniques are used for continuous variables?

a)

Binarization

b)

One-hot encoding

c)

Quantization

d)

Min-max scaling

e)

Bin counting

29.

Which of the following feature engineering techniques are used for categorical variables?

a)

Binarization

b)

One-hot encoding

c)

Quantization

d)

Min-max scaling

e)

Bin counting

30.

Which one of these two techniques is more scalable/efficient?

a)

One-hot encoding

b)

Feature hashing