Exam2

Question 1

Naive Bayes predicts…

Answer 1

What is the probability that a new data point has label A, B, C… etc

Question 2

Regression models are used to predict…

Answer 2

responses which have a continuous span of values

Question 3

The adjusted R^2 can be used

Answer 3

to compare models with different numbers of terms

Question 4

R^2, Mean Squared Error, Mean absolute error, etc, are examples of

Answer 4

goodness measures

Question 5

For Root Mean Squared Error (RMSE) if the value is < ? , it’s a good sign

Answer 5

< 1 standard deviation of the response variable

Question 6

For linear regression, always use…

Answer 6

more than 1 input variable

Question 7

Regression trees are

Answer 7

non-parametric method

Question 8

What is a stochastic process

Answer 8

a random variable that is a function of some index (ex space or time)

Question 9

What is Lazy Learning

Answer 9

A model fit using local data, it does not create a general model but instead memorizes the training data

Question 10

What is a technique that is an example of lazy learning

Answer 10

K nearest neighbors

Question 11

What is Eager Learning

Answer 11

A model fit which is “eager” to produce a general model to fit all data

Question 12

A Gaussian Process for regression produces

Answer 12

a probability distribution of functions

Question 13

For Gaussian Processes for regression, a Kernel function…

Answer 13

ascribes how each data is similar to others. It is chosen and determines the covariance function

Question 14

What are 3 main types of pre-processing for AI Model improvement

Answer 14

• Transformations
• Feature Selection
• Feature Engineering

Question 15

The selection of a subset of input variables to use in the model is called

Answer 15

Feature Selection (AI improvement)

Question 16

A “Transformation” is…

Answer 16

a mathematical alteration to the data space to facilitate improving the model goodness; rotating/stretching the input features

Question 17

PCA stands for ___ and is an example of

Answer 17

Principal Components Analysis, a Transformation technique for data pre-processing

Converts a set of correlated input variables to uncorrelated variables (ie Principal Components)

Question 18

What is “Feature Engineering”?

Answer 18

Addition of a new variable (feature) from existing variables to provide important info to ML, often adds human context to the data set (AI model improvement)

Question 19

What is a primary purpose of text analytics in model improvement?

Answer 19

To perform feature engineering to integrate text data into the ML process

Question 20

When you Transform the data using PCA, common application is

Answer 20

Reducing dimensions, to just use the most influential (largest) principal components, while minimizing information loss

Question 21

In PCA, the ____ are referred to as the principal components

Answer 21

Eigenvalues

Question 22

With PCA, the transformed space is…

Answer 22

Orthonormal and uncorrelated (0 covariance)

Question 23

Creation of Dummy Variables is considered a ___ technique

Answer 23

Transformation

Question 24

Dummy variables are…

Answer 24

a numerical equivalent to categorical variables

Question 25

If you have 3 or more categories to convert to a number, it is better to introduce a dummy variable than a scale to avoid...

Answer 25

introducing bias

Question 26

What is the limiting assumption of PCA

Answer 26

It is based on covariance, which is the linear statistical variation between 2 variables. The variables could have a more complex/nonlinear correlation in reality

Question 27

Covariance is...

Answer 27

the linear statistical variation between 2 variables

Question 28

What are some ML techniques that have built-in feature selection

Answer 28

Decision/Regression trees and Linear regression

Question 29

What is a disadvantage of feature selection

Answer 29

You're still losing information

Question 30

What are 3 types of importance measures used for feature selection

Answer 30

- Filter (select features in pre-processing, train on your selection) - Wrapper (train on a subset, then add or remove iteratively) - Embedded (integrated into learning process, ex Trees)

Question 31

What is the simplest importance measure to use for feature selection

Answer 31

Linear statistical importance - squared correlation

Question 32

"Feature Importance" means

Answer 32

assigning a numerical importance value to each feature

Question 33

When using squared correlation technique for feature importance, you could drop (input) features that...

Answer 33

show a very weak correlation to your target value (output) - ex 6.3

Question 34

Cross Validation can produce more _ models, while ensemble typically has the goal of more _ models

Answer 34

Robust, accurate

Question 35

One technique for Cross Validation is

Answer 35

K-folds

Question 36

_____ is a newer ensemble learning technique

Answer 36

Random Forests

Question 37

Ensemble learning often using ___ methods

Answer 37

decision/regression tree

Question 38

By themselves, Trees are considered ____

Answer 38

weak learners

Question 39

Trees can be prone to

Answer 39

overfitting

Question 40

Bagging can help

Answer 40

avoid overfitting

Question 41

Bagging definition in class

Answer 41

Each model uses a random subset of training data

Question 42

Boosting can help

Answer 42

remove bias

Question 43

What is the Random Forests technique

Answer 43

Ensemble method where a multitude of decision trees are used using probability measures for their construction and preduction

Question 44

The accuracy of what technique can get on same level as A NNs?

Answer 44

Random Forests

Question 45

CART stands for

Answer 45

classification and regression trees

Question 46

Bootstrap Aggregation is another term for

Answer 46

Bagging

Question 47

Text Analytics is the process of

Answer 47

quantifying information from raw text

Question 48

What are 3 methods of text analytics

Answer 48

- Feature-Value Mapping - Similarity Measures - Vectorizing

Question 49

Feature-value mapping is the same as...

Answer 49

Dummy Variables

Question 50

In text analytics, similarity methods work by...

Answer 50

calculating an equivalent distance metric for text variables

Question 51

What is an example of similarity method for text analytics

Answer 51

Levenshtein distance

Question 52

Definition of vectorizing

Answer 52

Conversion of raw text data into a numerical equivalent

Question 53

What are 3 ways to vectorize text?

Answer 53

- tokenizing - counting - normalizing

Question 54

What are "stop words"

Answer 54

words that may not be informative in a set of text data, that can be excluded from vectorization

Question 55

Validation test sets are used often when...

Answer 55

Detecting/avoiding overfitting when training ANNs