0% found this document useful (0 votes)
48 views

Assignment 1

The document discusses banking networks and how they can be represented as graphs. It provides examples of directed and undirected graphs as well as weighted and unweighted graphs. An example banking network with 7 accounts is presented and modeled as a graph. Properties of the graph like it being non-planar are discussed. Problems that could arise with large or changing graphs are also outlined. The conclusion discusses how machine learning is being applied in major US banks to improve customer service and operations through applications like natural language processing and predictive analytics.

Uploaded by

Bushra Kambo
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
48 views

Assignment 1

The document discusses banking networks and how they can be represented as graphs. It provides examples of directed and undirected graphs as well as weighted and unweighted graphs. An example banking network with 7 accounts is presented and modeled as a graph. Properties of the graph like it being non-planar are discussed. Problems that could arise with large or changing graphs are also outlined. The conclusion discusses how machine learning is being applied in major US banks to improve customer service and operations through applications like natural language processing and predictive analytics.

Uploaded by

Bushra Kambo
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

ASSIGNMENT – 1

TOPIC: BANKING NETWORK

MADE BY: BUSHRA KAMBO


Roll No: 6

COURSE: PGDDS 2017-2018


INSTITUTE: NISM & UNIVERSITY OF MUMBAI
Title: Banking Networks

Introduction:
To create a network we require the two most important aspects to
be taken into consideration namely as follows:
The Vertices & The Edges

Vertices are the nodes in a graph whereas edges are the lines
connected to these nodes.
Graphs can be undirected or directed:

Undirected graph:
The relationship exists in both directions.

Directed graph:
The relationships are based on the direction of the edges. It can be
a one way relationship or a two-way relationship, but it must be
explicitly stated.
In addition to this, graphs can be weighted or unweighted. All this
means is that there is some value or cost associated with the edges
between the vertices.
BANKING NETWORK GRAPH
The network consists of 7 bank accounts named as follows:
A/C1, A/C2,A/C3,A/C4,A/C5,A/C6,A/C7
Bank A/C 1 is connected to A/C2 & A/C7 (i.e A/C 7 is connected to
A/C1 & A/C2)
Bank A/C 2 is connected to A/C1,A/C7 & A/C3
Bank A/C3 is connected to A/C4 & A/C6 (Also to A/C2)
Bank A/C4 is connected to A/C3 & A/C5
Bank A/C5 is connected to A/C4 & A/C6
Bank A/C6 is connected to A/C3 & A/C5
The Graphical representation of the mentioned network is as
follows:
4
1

2 3 5

7
A 6
For the above Banking Network we can consider:
“Bank Accounts”  The vertices (Nodes/ ) , Denoted as “V”
“Transactions between two accounts”  The Edges (Connections/
), Denoted as “E”
“No of transactions between 2 accounts”  Weights, Denoted as
“Wt”
{A weighted graph is a graph where edges have positive real no’s
(non-negative)}

Properties of the above graph:


A graph with 5 or more vertices is not a PLANAR, thus the above
graph is a NON-PLANAR.

Weights (W) are assigned to the above graph:


[i.e Wt = {No of transactions between the two accounts}]
Wherein W(A)={2,3,7} and W(B)={1,4,5,6}
Therefore we get:
W(A,B) = 2+1+1+1=5
W(A)=2
W(B)=2
Applying optimisation formula:
Min(AB){W(A,B)/W(A)+W(A,B)/W(B)}
We get: 5/2+5/2=5
Optimisation value = 5
Higher the weight more the individuals are connected & vice
versa.
We can also have an unweighted graph.
If a graph is connected & there is no cycle – It is a “TREE” i.e if N-1
edges and no cycle, then it is a tree.
Forest ~ Collection of “TREES”

Problems to be faced:
Trouble starts as the graph becomes large.
Problem can occur when certain bank accounts are not connected
to each other directly.
Also problem with respect to changes in the details of a particular
bank account or any changes made to the original details of a
specific account, may disrupt connectivity.
In case if a bank account is closed it may again create a
connectivity error in the network.
Connectivity may be a problem at times because bank accounts
are highly secured and monitored, thus a single mistake in the
input of details for a particular account may lead to connectivity
issues.
Conclusion with respect to Banking Industry in reality:
Machine learning networks are seen as powerful tools to model
problems in order to extract information from data and for
prediction purposes.

The Federal Reserve of the US has recently published an


official report on the largest banks in the US. It lists quite a ton of
banks, yet we are not surprised by the fact 5 largest and most
influential banks of the US are investing heavily into imbuing their
services with Artificial Intelligence (AI) and Machine learning(ML).
Just to illustrate the efficiency of this approach — these banks have
closed more than 400 of local branches in 2016 and still met their
margin thresholds, as mobile banking combined with the ML helped
them meet and exceed their customer’s expectations.

The Machine Learning use cases are many : from sorting the email
using Natural Language Processing (NLP) and automatically
updating the records in the Customer Relations Management (CRM)
solution, to providing efficient assistance through customer self-
service portals and up to predicting the stock market trends in
order to ensure successful trading.

We will look through 5 use cases of machine learning in the banking


industry by highlighting the progress made by these 5 banks:
 JPMorgan Chase
 Wells Fargo
 Bank of America
 Citibank
 U.S. Bank
Bank show that this is the right direction and imbuing the banking
services with ML solutions is the way the industry should evolve in
the future.

You might also like