City Engineering College

Department of Computer Science and Engineering

Technical Seminar
on

"Machine Learning in Business Intelligence"

Under the guidance:

Prof. Vinod Kumar S
Name: Pratisha A S Karanth Asst. Proffesor, Dept
USN: 1CE18CS057 of CSE
TABLE OF
CONTENTS

• Introduction

• Approaches and Methods

• Conclusion and Future Enhancements

• References
 INTRODUCTION
• Business intelligence (BI) combines business analytics, data mining,
data visualization, data tools and infrastructure, and best practices
to help organizations to make more data-driven decisions.
• It is a comprehensive view of an organization’s data and that data
is used to drive change, eliminate inefficiencies, and quickly adapt
to market or supply changes.
• On the technical side, raw data is collected from the business’s
activity. Data is processed and then stored in data warehouses.
Once it’s stored, users can then access the data, starting the
analysis process to answer business questions
History of Business
Intelligence 1960s Sharing Information

Traditional Business Intelligence, originally

emerged in the 1960s as a system of sharing 1980s Decision Making
information across organizations. It further
developed in the 1980s alongside computer
models for decision-making and turning data
into insights before becoming specific 1990s Data Warehouses
offering from BI teams with IT-reliant service
solutions. Modern BI solutions prioritize
flexible self-service analysis, governed data
on trusted platforms, empowered business
users, and speed to insight.
IMPORTANCE OF BUSINESS
INTELLIGENCE
01
Gain New Customer Insights: One of the primary
reasons companies are investing their time, 02
money, and efforts into Business Intelligence is Improved Visibility: Business Intelligent organizations
because it gives them a greater ability to have better control over their processes and
observe and analyze current customer buying standard operating procedures, as the visibility of
trends. these functions is improved by a BI system.

03
Efficiency Improvements: BI Systems help improve
organizational efficiency which consequently increases 04
productivity and can potentially increase revenue.
Competitive Advantage: In addition to all of these great
Business Intelligence systems allow businesses to share
benefits, Business Intelligence can help you gain
vital information across departments with ease, saving
insight into what your competitors are doing, allowing
time on reporting, data extraction, and data
your organization to make educated decisions and
interpretation.
plan for future endeavors.
MACHINE LEARNING IN BUSINESS
INTELLIGENCE
• One of the key limitations of BI tools has been that, whilst they are effective for
viewing, exploring, and interacting with data, they are not very effective or
efficient at uncovering hidden insight.
• Machine learning models are very effective at uncovering hidden patterns and
insight in data.
• Of late, we are starting to see a number of BI vendors incorporating machine
learning capabilities into BI tools, bringing with it the promise of making BI much
more effective at identifying hidden insights.
MACHINE LEARNING IN BUSINESS
INTELLIGENCE

• In addition to finding these hidden insights, the automation of this process means
that insights can be delivered much faster, thus allowing the business to act quickly
and with better information.
• Many analysts are performing routine tasks such as variance analysis and looking
for anomalies. If these tasks are automated, the analyst can be freed up to work on
higher-value tasks.
IMACHINE LEARNING IN BUSINESS
INTELLIGENCE
LITERATURE SURVEY
Serial Title Authors Year Methodology
No.
1 Improving Business Shaista Anwar, 2021 Research methodology used during the study
Intelligence through Vinod Kumar conducted in this research paper is based
Machine Shukla, Minsang D. both upon primary as well as secondary data.
Learning Algorithms Tamang A few most commonly used machine learning
algorithms which could be best used,
according to their needs and challenges, have
also been
2 Business Intelligence Ruchi Sharma, Dr 2018 discussed
Previously,inoutliers
the paper.
were being considered as
using Machine Pravin Srinath noisy data and disregarded leading to loss of
Learning and Data relevant information. This paper highlights
Mining techniques-An the major research challenges in this mining
analysis sub domain. It provides a comprehensive
taxonomy extracted for Business Intelligence
methodologies along with current application
sectors.
 LITERATURE SURVEY
3 Effective Muhammad 2020 Demand forecasting is one of the main decision-making tasks of
Demand Adnan Khan, enterprise. The machine learning engine executes data from
Forecasting Shaiza Saqib, different modules and determines the weekly, monthly, and
Model using Tahir Alyas, quarterly demands of goods/commodities. In demand forecasting,
Business Anees Ur its perfect accuracy is non-compromising, the more accurate
Intelligence Rehman system model is more efficient. Furthermore, we test the efficiency
empowered by comparing the predicted data with actual data and determine
with Machine the percentage error. Simulation results show that after applying
Learning the purposed solution on real-time organization data, we get up to
92.38 % accuracies for the store in terms of intelligent demand
4 Discovering Liping Zhao, 2021 The Metis+ framework, intended to discover business
forecasting.
Business Sam Supakkul, problems using a goal-oriented and ML approach, provides
Problems Robert Ahn, an ontology (or modeling concepts), semantic reasoning
Using Problem Ronaldo methods, and processes.
Hypotheses: A Goncalves
Goal-Oriented Junior, Tom Hill,
and Machine Lawrence
Learning- Chung
Based
PROBLEM STATEMENT

Enterprises need to analyze a lot more than just profit or loss.

To uncover hidden patterns in this omnipresent data and gain
knowledge for business intelligence is becoming extremely
crucial.
EXISTING SYSTEM
In the existing system, Business intelligence encompasses a set of
analytics techniques to transform the raw data into meaningful and useful
information. Business intelligence tools enable to extract relevant data,
make it reliable, accurate and attractive to users.

PROPOSED SYSTEM
Explosion of data is making it more and more challenging to extract
meaningful information and gather knowledge by using standard
algorithms, due to the increasing complexity of analysis. There is a
growing understanding that business intelligence (BI) and machine
learning (ML) are facilitators of each other and perform best in
conjunction, for large volumes of data being utilized for business needs.
APPROACHES AND METHODS

• Combining machine learning with BI can have a far-reaching impact on the

insights a business gets from its available data.
• Beyond demand models that only predict market trends, revenue levels and
so on, ML can generate actual answers. As some businesses are now
discovering, ML lets them produce highly accurate estimates of future
behavior—i.e., answers—based on large volumes of historic data.
• Algorithms like K-Nearest Neighbors, Random Forest, Linear Regression can
be used to predict the sales and other required numbers.
APPROACHES AND METHODS
• BI processes involve analyzing large sets of data—if done manually, this is
time-consuming. Machine learning can automate the process so that BI
professionals can focus on higher-level trend analysis and behavior patterns
that can bring greater value more quickly to the organization.
• Through machine learning, anomalies can be spotted in real time and
immediate action can be taken. For instance, fraud can be spotted right away
—not a week later—or customers can be kept on your website rather than
learning about them after they have bought something elsewhere.
• Systems can be created immediately to avoid the anomalies in the future,
increasing operational efficiency.
APPROACHES AND METHODS

• ML can translate data concepts into business

language and use that information to establish a
data warehouse cloud. This can act as a semantic
layer that maps BI concepts within a data
architecture, creating a credible and good-quality
point of reference for the business.
• This will build a digital environment that
understands your business and is better suited to
answer specific questions.
APPROACHES AND METHODS

Case Study-1: BI, ML IN AIRPORT SYSTEM

The following discussion is based on “Airport
System”. Business Intelligence practices at an
airport: As it is known to everyone, an airport is a
place where a lot of activities take place at the
same time.
APPROACHES AND METHODS

• The business intelligence tools used by airport system are the following:
• Oracle Business Intelligence Enterprise Edition or OBIEE is the set of
business intelligence tools of Oracle Corporation, which also consists of
former Siebel Systems and Hyperion Solutions business intelligence
offerings. The architecture of OBIEE consists of the clients, oracle BI
presentation services, oracle BI repository, data sources and the Oracle BI
server which is the central component.
APPROACHES AND METHODS
• Oracle BI Publisher - This is an enterprise reporting solution whose primary
feature is separating the data extraction process from the reporting layout.
• Oracle data warehouse to extract BI reporting – Data warehousing is
considered a key component in business intelligence. It is used for reporting
and storing data analytics.
According to the requirements in this case,
• Logistic regression algorithm can be used. This is a supervised machine
learning algorithm based on the concept of probability and is a good option
because of its easier implementation and interpretation. The outputs are
numbers between 0 and 1 that can be interpreted as a probability.
APPROACHES AND METHODS

• Forecasting sales is an important field in any form of business organization. It

lets us know about which factors play a major role in gaining profit and what
factors to be avoided for the same.
• Algorithms like simple linear regression and random forest regression can be
used for making predictions.
CONCLUSION AND FUTURE
ENHANCEMENTS
• Implementing machine learning in business intelligence is going to prove to be
much more effective and powerful, if the right algorithms are used according to
the needs of the business.
• With increasing advancements in technology, artificial intelligence will continue to
grow which gives businesses more chance to integrate the insights into a broader
business strategy.
• Hence, it can be said that machine learning is going to be an integral part of all
systems, both small as well as large, as we already know that artificial intelligence
has already made changes in major industry sectors such as manufacturing,
healthcare, media and retail
CONCLUSION AND FUTURE
ENHANCEMENTS

• There is definitely going to be a strong dependence on Business Intelligence to

advance user acquisition and growth in every sector.
• As the dependence on data increases, almost any usage of data will require
analytics and Business Intelligence to extract actionable insights in future.
• The Business Intelligence Models can be made more accessible to the small
business owners.
• The models can be made more adavance and be incorporated to self-fix the
minor problem areas.
 REFERENCES

[1] “Improving Business Intelligence through Machine Learning Algorithms” 2nd

International Conference on Intelligent Engineering and Management (ICIEM), IEEE
Xplore
[2] https://www.tableau.com/learn/articles/business-intelligence
[3] https://www.dataversity.net/brief-history-business-intelligence/
[4] https://www.yellowfinbi.com/blog/2018/02/impact-of-machine-learning-on-
business-intelligence
REFERENCES

[5] https://www.datanami.com/2020/02/13/ml-and-bi-are-coming-together-gartner-
says/
[6] https://www.forbes.com/sites/forbestechcouncil/2019/10/30/12-impactful-ways-
to-incorporate-machine-learning-into-business-intelligence/?sh=626413686194
[7] https://link.springer.com/article/10.1007/s11042-022-13132-3
Thank You