Chapter 1: Machine Learning for Predictive Analysis

Question 1

What is the job of data analytics?

Answer 1

Extracting insights from data

Question 2

What is predictive data analytics?

Answer 2

The art of building and using models that make predictions based on patterns extracted from historical data

Question 3

What are the applications of predictive data analysis?

Answer 3

price prediction (businesses), dosage prediction (doctors), risk assessment (organizations), propensity modeling (predicting the likelihood or propensity of individuals to take different actions), diagnosis (doctors, engineers and scientists), document classification

Question 4

What is a prediction in data analytics? How is it different from the everyday usage?

Answer 4

In DA a prediction is the assignment of a value to any unknown variable. In everyday usage it has a temporal aspect, we predict what will happen in the future

Question 5

What two things are common in all the application examples?

Answer 5

in each case, a model is used to make a prediction to help make a decision AND a model is trained to make predictions based on a set of historical examples (machine learning is used to train these models)

Question 6

What is machine learning?

Answer 6

Machine learning is an automated process that extracts patterns from data.

Question 7

What is supervised machine learning used for?

Answer 7

We use supervised machine learning to build the models used in predictive data analytics applications
- They have labels/ classes/ events that provide us with feedback while learning

Question 8

How do they work?

Answer 8

They automatically learn a model of the relationship between a set of descriptive features and a target feature based on a set of historical examples (instances)

Question 9

What is each row of a dataset called?

Answer 9

training instance

Question 10

What is the overall dataset called?

Answer 10

training dataset

Question 11

When is a model consistent?

Answer 11

when there are no instances in the dataset for which the model does not make a correct prediction

Question 12

What do machine learning algorithms do?

Answer 12

automate the process of learning a model that captures
the relationship between the descriptive features and the target feature in a dataset.

Question 13

Why is searching for consistent models not enough to learn useful prediction models?

Answer 13

• When dealing with large databases there will likely be noise
• The training set represents only a small sample of the possible set of instances in the domain.

Question 14

What is an ill-posed problem?

Answer 14

An ill-posed problem is a problem for which a
unique solution cannot be determined using only the information that is available

Question 15

Why is machine learning an ill-posed problem?

Answer 15

A single consistent model cannot be found based on the sample training dataset alone

Question 16

What is generalization?

Answer 16

• The ability to make predictions for queries that are not present in the data
• A prediction model that makes the correct predictions for these
  queries captures the underlying relationship between the descriptive and target features
  and is said to generalize well

Question 17

What is the goal of machine learning?

Answer 17

Finding the predictive model that generalizes best

Question 18

What is inductive bias?

Answer 18

A set of assumptions that defines the model selection criteria of a machine learning algorithm

Question 19

What are the types of inductive bias?

Answer 19

• Restriction bias
  -Preference bias

Question 20

What is restriction bias?

Answer 20

• It constrains the set of models that the algorithm will
  consider during the learning process
• Similar to choosing your go-to study method
• Tells us what our model is able to represent

Question 21

What is preference bias?

Answer 21

-It guides the learning algorithm to
prefer certain models over others
- Choosing out convergence/satisfaction mechanism
- I like group study but prefer to be the leader than weakest link
- Algorithm’s belief about what makes a good hypothesis

Question 22

What are examples of restriction bias?-

Answer 22

• In multivariable linear regression with gradient descent we only consider models that produce description based on a linear combination of the descriptive features
• In Iterative Dichotomizer 3 we only consider tree-like prediction models where each branch encodes a sequence of checks on individual descriptive features

Question 23

What are examples of preference bias?

Answer 23

• In MLR with GD we linearly combine the descriptive features using only weights that were found though our gradient descent approach
• In ID3 we are preferring shallower (less complex) trees over larger/deeper trees

Question 24

Why is inductive bias necessary for learning beyond the dataset?

Answer 24

Without it we could only perform memorization of our training dataset without generalization capacity

Question 25

What is model induction?

Answer 25

The creation of models from data

Question 26

What is the difference between classification problem and regression problem?

Answer 26

- Classification problem has the target as a category - Regression problem has the target as a number

Question 27

What is another name for dataset?

Answer 27

-One whose form is the same is a table/relation of a database - Worksheet of a spreadsheet - Array in math

Question 28

What is an instance

Answer 28

- Row, tuple, record of a database table - Case in statistics - Object of a class in programming - Datapoint or vector in math

Question 29

What is an independent variable?

Answer 29

- It is the attribute supplied as input - Also known as explanatory variable, inputs, predictors -Features are the table's columns

Question 30

What is dependent variable

Answer 30

The target variable whose values are to be predicted. Aka class or label or output

Question 31

What are some confusing facts about independent and dependent variables?

Answer 31

- Independent variables may not be independent on each other or anything else - Dependent variables does not always depend of all the independent variables

Question 32

Facts about the target variable

Answer 32

- Sometimes it is considered to be included in the set of features, sometimes it is not - The target variable is not used to predict itself - Prior values may be helpful to predict future values and may be included as input features

Question 33

What is the process of building a model (or training your classifier) from historical data?

Answer 33

Induction, learning, training or generalization

Question 34

When does the real value of machine learning become apparent?

Answer 34

When we want to build prediction models from large datasets with multiple features

Question 35

How do you know the number of possible prediction models?

Answer 35

- There are three descriptive features so there are 2^3 possible combinations of descriptive feature values - For each descriptive feature there are 3 possible target feature values - There are 3^8 = 6,561 possible prediction models

Question 36

What is the ability to memorize a training dataset?

Answer 36

Consistency

Question 37

What does Occam's Razor say about simplicity?

Answer 37

With all things being equal the simplest explanation tends to be the right one (upper bound)

Question 38

What does albert Einstein say about simplicity

Answer 38

Everything should be made as simple as possible but not simpler (lower bound)

Question 39

What are the sources of information that guide machine learning algorithms?

Answer 39

- Training data - Inductive bias of the algorithm

Question 40

What can go wrong with machine learning?

Answer 40

- Inappropriate inductive bias which leads to mistakes

Question 41

What does no free lunch mean?

Answer 41

if an algorithm does well on a certain class of problems then it necessarily pays for that with degraded performance on the set of all remaining problems

Question 42

What happens if we choose the wrong inductive bias?

Answer 42

- Underfitting (the prediction model is oversimplifies) -Overfitting (the prediction model is so complex that it becomes too sensitive to noise in the data, it memorizes)

Question 43

What is a Goldilocks model?

Answer 43

- A model that is just right and strikes a good balance between overfitting and underfitting - it is found by using algorithms with appropriate inductive biases

Question 44

What is CRISP-DM

Answer 44

Cross Industry Standard Process for Data Mining is a data mining process model that describes commonly used approaches that data mining experts use to tackle problems

Question 45

What are the phases?

Answer 45

-Business Understanding- defining customers' needs, understanding project objectives -Data Understanding- collection and data familiarity -Data Preparation- construct final dataset from raw data -Modeling- select machine learning techniques relevant to the problems and their parameters are calibrated to optimal values -Evaluation- outcome collection, compare obtained model with business objectives -Deployment- put into production, organize and present knowledge gain in a way that the customer can use it

Question 46

List other data life cycle models

Answer 46

-Semma- Sample, Explore, Modify, Model, Assess - Data Mining and Knowledge Discovery from Data (KDD)- mostly used in the real world

Question 47

What is supervised machine learning based on?

Answer 47

The assumption that data does not change over time. They create models that distinguish between classes present in the dataset they are induced from