B07 Logistic Regression

Question 1

Regression

Machine learning technique where the goal is to model
the ____ and ____ of numeric relationships in order
to predict a target variable based on the values of
previously observed explanatory variables.

Answer 1

Size and Strength

Question 2

Popular regression techniques include:

Answer 2

Linear Regression
Logistic Regression
Poisson Regression (log-linear)

Question 3

Linear Regression:

The goal is to estimate the values for ___ and __ in order
to minimize _____________________________
between the actual and predicted values.

Answer 3

B0 and B1 (Beta)

The sum of squared errors

Question 4

Limits of Linear Regression

Answer 4

• requires a normal distribution of the response variable

- decision boundary

Question 5

In logistic regression, the coefficients are estimated

using a technique called __________________

Answer 5

Maximum Likelihood

Estimation (MLE)

Question 6

Unlike the _________________ method
used by linear regression, finding a closed form for
the coefficients using MLE is not possible. Instead, the
process is iterative.

Answer 6

Ordinary Least Squares (OLS)

Question 7

The ___________________________is an extension of
linear regression that allows for linear predictors to be
related to a response variable that is not normally
distributed by using a transformation or link function

Answer 7

Generalized Linear Model (GLM)

Question 8

The link function used for binomial logistic regression is called the _________

Answer 8

logit function

Question 9

In Logistic Regression

For every unit increase in
tumor size, the odds of it
being malignant changes
by a multiple of ___

Answer 9

e^Beta

Question 10

If B<0 , the odds
that the tumor is
malignant \_\_\_\_\_\_\_ as
tumor size increases.
If  B>0, the odds
that the tumor is
malignant \_\_\_\_\_\_\_\_ as
tumor size increases.

Answer 10

decreases

increases

Question 11

An estimate of the relative information lost by a given
model: the less information a model loses, the higher
the quality of the model

Answer 11

Akaike Information Criterion (AIC)

Question 12

Strengths of Logistic Regression?

Answer 12

-Outputs have a nice probabilistic
interpretation.
-Can be regularized to avoid
overfitting.
-Easy to implement and use.
-Very efficient to train.

Question 13

Weaknesses of Logistic Regression?

Answer 13

-Makes strong assumptions
about the data.
-Does not do well with missing
data.
-Tends to underperform when
there are multiple or non-linear
decision boundaries.
-Does not naturally capture
complex relationships.