Terminology Week 1

Question 1

Describe linear regression.

Answer 1

A technique to find a relation between one or many input variables and an output variable.
A technique to find a line to best fit a set of points.

Question 2

Describe a linear function?

Answer 2

y = w*x + b
y = predicted value (predicated label)
x = input variable (feature)
w = weight (weight vector, gives the slope)
b = initial bias

Question 3

Describe the notion of loss.

Answer 3

How well does our line predict an example.
How far away is the data point from the line.
Loss is a number indicating how bad the model’s prediction was on a single example.

We can describe the loss as the difference between the prediction of the value (a dot on the line) and the actual value (the example).

Question 4

How can you call the L2 loss function?

Answer 4

Also called squared loss.

Question 5

How can you call the squared loss function?

Answer 5

L2 loss function

Question 6

Describe the L2 loss function? And how is it calculated for a data set?

Answer 6

It is the square of the difference between the predicated value (predication) and the true value (label).

We don’t care about the loss of a single value. We want to minimise the loss across our entire data set.

Sum of all squared differences (predicated - example)

Often devided by the number of examples in the data s = average square loss = Mean square error = MSR

Question 7

What doe Infer mean?

Answer 7

To predict

Question 8

What is linear regression (google definition)?

Answer 8

A type of regression model that outputs a continuous value from a linear combination of input features.

Question 9

What is Inference (google definition)?

Answer 9

In machine learning, often refers to the process of making predictions by applying the trained model to unlabeled examples. In statistics, inference refers to the process of fitting the parameters of a distribution conditioned on some observed data.

Question 10

What does training a model mean?

Answer 10

Training a model simply means learning (determining) good values for all the weights and the bias from labeled examples. In supervised learning, a machine learning algorithm builds a model by examining many examples and attempting to find a model that minimizes loss; this process is called empirical risk minimization.

Question 11

Describe MSE.

Answer 11

Mean square error = average squared loss