1) Decision tree

Decision tree methodology

is a usually used data mining method for founding classification systems based

on multiple covariates or for evolving prediction algorithms for a target

variable.

The basic concept of the

decision tree

1.

Nodes. There

are three types of nodes. (Lu and Song, 2017)

–

A root hub, likewise called a choice hub, speaks

to a decision that will bring about the subdivision of all records into at

least two totally unrelated subsets.

–

Internal hubs, likewise called possibility hubs,

speak to one of the conceivable decisions accessible by then in the tree

structure, the best edge of the hub is associated with its parent hub and the

base edge is associated with its kid hubs or leaf hubs.

–

Leaf hubs, additionally called end hubs, speak to

the last aftereffect of a blend of choices or occasions.

2. Branches. (Lu and Song, 2017)

–

Branches

represent chance outcomes or occurrences that emanate from root nodes and

internal nodes.

–

A decision

tree model is formed using a hierarchy of branches. Each path from the root

node through internal nodes to a leaf node represents a classification decision

rule.

–

These

decision tree pathways can also be represented as ‘if-then’ rules.

3.

Splitting. (Lu and Song, 2017)

–

Only input

variables related to the target variable are used to split parent nodes into

purer child nodes of the target variable.

–

Both discrete input

variables and continuous input variables which are collapsed into two or more

categories can be used.

–

When building the

model one must first identify the most important input variables, and then

split records at the root node and at subsequent internal nodes into two or

more categories or ‘bins’ based on the status of these variables.

The type of the decision tree

·

Classification tree analysis is when the

predicted outcome is the class to which the data belongs.

·

Regression tree analysis is when the

predicted outcome can be considered a real number (e.g. the price of a house,

or a patient’s length of stay in a hospital).

2. Logistic Regression

–

Logistic regression is used to find the

probability of event=Success and event=Failure. We should use logistic

regression when the dependent variable is binary (0/ 1, True/ False, Yes/ No)

in nature.

–

The binary

logistic model is charity to estimate the probability of a binary response based

on one or more predictor (or independent) variables (features).

–

It allows

one to say that the presence of a risk factor increases the odds of a given

outcome by a specific factor.

–

Logistic regression doesn’t require

linear relationship between dependent and independent variables. It can handle various types of relationships

because it applies a non-linear log transformation to the predicted odds ratio.

(Sachan,2017).

The type of logistic regression

1. Binary logistic regression (Wiley,2011)

–

used when the dependent variable is

dichotomous and the independent variables are either continuous or categorical.

–

When the

dependent variable is not dichotomous and is comprised of more than two

categories, a multinomial logistic regression.

2. Multinomial Logistic Regression (Wiley,2011)

–

The linear

regression analysis to conduct when the dependent variable is nominal with more

than two levels. Thus it is an extension of logistic regression, which analyses

dichotomous (binary) dependents.

–

Multinomial

regression is used to describe data and to explain the relationship between one

dependent nominal variable and one or more continuous-level (interval or ratio

scale) independent variables.

The logistic regression does not assume a linear relationship between

the independent variable and dependent variable and it may handle nonlinear

effects. The dependent variable need not be normally distributed. It does not

require that the independents be interval and unbounded. Logistic regression

come at a cost, it requires much more data to achieve stable, meaningful

results. logistic regression come at a cost: it requires much more data to

achieve stable, meaningful results. With standard regression, and dependent

variable, typically 20 data points per predictor is considered the lower bound.

For logistic regression, at least 50 data points per predictor is necessary to

achieve stable results (Wiley,2011)

3) Neural Network

Neural network is a method of the computing,

based on the interaction of multiple connected processing elements. Ability to

deal with incomplete information. When an element of the neural network fails,

it can continue without any problem by their parallel nature.

(Liu, Yang and Ramsay, 2011)

Basic concept of the

neural network (Liu, Yang and Ramsay, 2011)

1.

Computational Neuroscience

–

understanding and modelling operations of

single neurons or small neuronal circuits, e.g. minicolumns.

–

Modelling information processing in actual

brain systems, e.g. auditory tract.

–

Modelling human perception and cognition.

2.

Artificial Neural Networks

–

Used in Pattern recognition, adaptive

control, time series prediction and etc.

–

The

areas contributing to Artificial neural networks are Statistical Pattern

recognition, Computational Learning Theory, Computational Neuroscience,

Dynamical systems theory and Nonlinear optimisation.

The type of neural

network (Hinton,2010)

1.

Feed-Forward neural network

–

There is the commonest type of neural

network in practical application. The first layer is the input and the last

layer is output.

–

If the is more than one hidden layer, we

call them ‘deep’ neural networks. They compute a series of transformation that

change the similarities between cases.

2.

Recurrent networks

–

These have directed cycles in their

connection graph. That means you can sometimes get back to where you started by

following the arrows.

–

They can have complicated dynamic and this can

make them very difficult to train.

A neural network can perform tasks that a linear program cannot. A

neural network learns and does not need to be reprogrammed. It can be

implemented in any application. It can be implemented without any problem. Neural

networks requiring less formal statistical training, ability to implicitly

detect complex nonlinear relationships between dependent and independent

variables, ability to detect all possible interactions between predictor

variables, and the availability of multiple training algorithms. (JV,1996)

Factors

Decision Tree

Logistic

Regression

Neural Network

Basic concept

1.

Nodes: Root node, Internal node, Leaf nodes

2.

Branches

3.

Splitting

output can take

only two values, “0” and “1”, which represent outcomes

such as pass/fail and win/lose

1. Computational

Neuroscience

2.Artificial Neural

Networks

Type

1.

Classification tree analysis

2.

Regression tree analysis

1. Binary

logistic regression

2. Multinomial

logistic

regression

1.

Feed-forward network

2.

Recurrent decnetwork

Performance

·

Can quickly express complex alternatives

clearly

·

Can easily modify

·

Standard decision tree notation is easy to adopt.

·

It does not assume a linear relationship between the

independent variable and dependent variable

·

It may handle nonlinear effects

·

Requires much more data to achieve stable

·

Can perform tasks that a linear program cannot.

·

Requiring less formal statistical training

·

Ability to implicitly detect complex nonlinear

relationships

·

Ability to detect all

possible interactions