Neural Networks Intro

Question 1

What does creating a machine learning algorithm mean?

Answer 1

It means building a model that outputs correct information from provided input data

Question 2

What is the training process?

Answer 2

The process which the model learns to make sense of the input data.

Question 3

What are the 4 ingredients of training an algorithm?

Answer 3

1. Data
2. Model
3. Objective Function
4. Optimization Algorithm

Question 4

Which of the following is NOT a building block of a machine learning algorithm?

Data

Variable

Objective function

Optimization algorithm

Answer 4

Variable

Question 5

Training the model is:

A pure trial and error process

A kind of trial-and-error process with some feedback

The process of giving guidelines to the computer to find patterns

The process of watching people or a machine perform an activity and replicating it

Answer 5

A kind of trial-and-error process with some feedback

Question 6

Self-driving cars learn by:

Driving many hours before learning how to do it safely and efficiently

A very strict set of rules that Elon Musk and the others are programming day and night

“Watching” thousands of hours of footage of real people driving

Breaking the rules (e.g., go on red light, go over the speed limit) and waiting to get punished for it

Answer 6

“Watching” thousands of hours of footage of real people driving

Question 7

What are the 3 types of machine learning?

Answer 7

1. Supervised
2. Unsupervised
3. Reinforcement

Question 8

What are the two subtypes of supervised learning?

Answer 8

1. Regression
2. Classification

Question 9

The linear model is given by:

y = xT w + b

y = wx + b

y = wT x + b

All of the above

Answer 9

All of the above

Question 10

The linear model for multiple inputs is given by:

y = xw + b

y = x1w1 + x2w2 + b

y = x1w2 + x2w1 + b

y = x2w2 + x2w2 + b

Answer 10

y = xw + b

Question 11

You have y = xw + b, where w = [1.2, -3], while b = [7]. If x = [2 , 3], what is the value of y?

0.4

-13.6

4.6

9.4

Answer 11

0.4

Question 12

What is the Objective Function?

Answer 12

the measure used to evaluate how well the model’s outputs match the desired correct values

Question 13

What are the two Objective Function types?

Answer 13

Loss - minimizing the error

Reward - reinforcement earning - super mario hight score example

Question 14

In supervised learning, we are dealing with:

lost functions

loss functions

reward functions

reinforcement functions

Answer 14

loss functions

Question 15

Reward functions are NOT:

functions we are trying to maximize

functions used in reinforcement learning

functions we are trying to minimize

functions

Answer 15

functions we are trying to minimize

Question 16

What is the common Loss Function dealing with Regresssion?

Answer 16

L2-norm

Question 17

What is the common Loss Function when dealing with Classificaiton

Answer 17

Cross-entropy

Question 18

What is a Target?

Answer 18

The desired outcome. We want Y to be as close to T as possible

Question 19

What are the outputs of a regression?

Answer 19

continuous numbers

Question 20

A target is:

The correct value at which we are aiming

A synonym for output

A part of the model

Always bigger than 0

Answer 20

The correct value at which we are aiming

Question 21

The objective function measures:

how well the targets match our model’s outputs

how well our model’s outputs match the targets

the model’s parameters

linearity of the data

Answer 21

how well our model’s outputs match the targets

Question 22

The L2-norm loss is used for:

k-means clustering

classification

regression

hierarchical clustering

Answer 22

regression

Question 23

Cross-entropy loss is used for:

k-means clustering

classification

regression

hierarchical clustering

Answer 23

classification

Question 24

Which cross-entropy points to the best match between outputs and targets?

L = 12.41

L = 0.78

L = 0.44

L = 0.77

Answer 24

L = 0.44

Question 25

The cross-entropy loss divided by 10 cannot be used for machine learning

True

False

Answer 25

False

Question 26

The cross-entropy loss MULTIPLIED by 10 cannot be used for machine learning

True

False

Answer 26

False

Question 27

The gradient is:

a generalization of the derivative concept

a generalization of the integral concept

a generalization of the optimization algorithm

a generalization of the objective function

Answer 27

a generalization of the derivative concept

Question 28

The gradient descent is a type of:

data

model

objective function

optimization algorithm

Answer 28

optimization algorithm

Question 29

The learning rate is denoted by which Greek letter?

alpha

sigma

nabla

eta

Answer 29

eta

Question 30

A high learning rate:

is better, as you learn faster

is faster, but may not reach the minimum

is slower, but sure to reach the minimum

is faster, and always reaches the minimum

Answer 30

is faster, but may not reach the minimum

Question 31

How is the Loss Function denoted?

Answer 31

L(y,t) - loss
C(y,t) - cost
E(y,t) - error

Question 32

N-parameter gradient descent differs from the 1-parameter gradient descent as it deals with:

many weights and biases

many input variables

many output variables

many targets

Answer 32

many weights and biases

Question 33

We use the delta to denote:

difference in models

difference between outputs and inputs

difference between outputs and targets

difference between methodologies

Answer 33

difference between outputs and targets

Question 34

The weights and biases:

Are the same thing

Have different update rules

Are rarely updated

Have update rules following the same logic

Answer 34

Have update rules following the same logic

Question 35

What determines a model uniquely?

Answer 35

The weights (w), biases (b), and model architecture

Question 36

Training in a deep network is done by adjusting the parameters in the direction of minimizing the loss. How do you find that direction?

Answer 36

By calculating the gradient of the loss with respect to the weights

Question 37

Why is the loss sometimes divided by the number of observations/data points?

Answer 37

to keep the same learning rate for different sized datasets