Machine learning

Question 1

What are the variable types? Draw the diagram

Answer 1

Discrete Continuous
/ \ |
Categorical Numerical Numerical
/ \ | |
Nominal Ordinal Interval Ratio

Question 2

Properties of attribute types

Answer 2

Nominal Ordinal Interval Ratio
Distinct ✅ ✅ ✅ ✅
Order ❌ ✅ ✅ ✅
Addition ❌ ❌ ✅ ✅
Multiplication ❌ ❌ ❌ ✅

Question 3

What is global standardisation?

Answer 3

Feature scaling to ensure all features have the same mean and SD especially for larger features that dominate smaller features

Question 4

What is dimensionality reduction?

Answer 4

A technique to reduce the number of features in a dataset while preserving the relevant information (PCA)

Question 5

What is feature selection?

Answer 5

Selects a subset of original features that are the most relevant which acts as feature removal. Reduces dimensionality of dataset

Question 6

What is feature extraction?

Answer 6

Identification of a reduced set of transformed features which contributes to a reduction of dimensionality in the dataset (PCA)

Question 7

Benefits of feature selection

Answer 7

Simplifies the model, reduces size of dataset, improves model accuracy, more efficient in training and easier to interpret

Question 8

Most important things to check for in data cleaning

Answer 8

Outliers, missing values and duplicates

Question 9

Benefits of sampling

Answer 9

It allows for quicker analysis when analysis on whole dataset is not feasible

Question 10

Types of sampling

Answer 10

Subsampling, sampling, re sampling, random sampling and stratified random sampling

Question 11

What is subsampling

Answer 11

Used for data reduction by selecting a subset of original dataset

Question 12

What is sampling?

Answer 12

Creation of training and testing datasets

Question 13

What is resampling?

Answer 13

Repeatedly drawing samples to estimate the characteristics of the whole dataset. Used for bias removal (bootstrapping)

Question 14

What is random sampling?

Answer 14

Without replacement (pick balls out the bag)
With replacement (picking balls but putting them back)

Question 15

What is stratified random sampling?

Answer 15

Random samples are taken from each variable based on relevant features. Used to contribute to model performance

Question 16

What is the holdout method?

Answer 16

Splitting the dataset into 2 parts and holdout part of the dataset and use the other part to train (test and train)

Question 17

What is k fold cross validation?

Answer 17

Evaluates model performance by dividing the dataset into k number of folds. Once one is used, the process is repeated k times with a different combination each timw

Question 18

What is bootstrapping?

Answer 18

Create repetitions of the sets using random sampling generally > 1000 which is used to reduce bias and make dataset more robust

Question 19

Types of model performance evaluation

Answer 19

Regression, classification and binary classification

Question 20

How do you do regression performance evaluation?

Answer 20

Mean squared error, root mean square and mean absolute error

Question 21

How do you do classification performance evaluation?

Answer 21

Actual class value vs predicted class value, confidence on prediction and confusion matrix

Question 22

How do you do binary classification performance evaluation?

Answer 22

Using class of interest instances falling into positive for the class and negative for the other classes

Question 23

How to evaluate model performance when there’s a class imbalance for both regression and classification?

Answer 23

Kappa statistic
Roc curve

Question 24

What does the kappa statistic do?

Answer 24

It adjusts accuracy for correct predictions by chance and partially mitigates the effects of class imbalance

Question 25

K nearest neighbour benefits and disadvantages

Answer 25

Simple and easily explained and interpreted
Fast in training
Both classification and regression

Doesn’t produce a model
Slow in classification
Difficult to select k

Question 26

SVM benefits and disadvantages

Answer 26

High accuracy in prediction
Compact model representation
Not prone to overfishing

Difficult to select kernel/parameters
Slow to train for large datasets
Not easily explained

Question 27

Decision trees benefits and disadvantages

Answer 27

Human readable structural pattern
Transparent model for numeric and nominal features
Easily explained and interpreted

Inefficient for high dimensional data
Tendencies to overfit
Sensitive to small perturbations

Question 28

Random forests benefits and disadvantages

Answer 28

Efficient for high dimensional data
Insensitive to noise/missing data
Directly selects important features for both numerical/nominal features

Not easily explained or interpreted
More difficult to tune
Biased for features with many levels

Question 29

Neural networks benefits and disadvantages

Answer 29

Model complex patterns in data
Both classification and regression
No assumption of data relationships

Not easily interpreted
Prone to overfitting in training
Computationally intense to train

Question 30

Centre based clustering k means benefits and disadvantages

Answer 30

Efficient (linear complexity
Automatic and no cut off required

Depends on intial seeds
Often trapped in local minima
Inefficient for high dimensional data