INTRODUCTION TO BUSINESS INTELLIGENCE and DATA MINING

Preview:

DESCRIPTION

 

Citation preview

Rajesh Math

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 111

BI I Lecture 1

Syllabus/Course

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 122

Same group for communicationSyllabusBooksTools/SoftwareTitle of Course is BI

Data Mining

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 133

Data Mining is process of exploration and analysis by automatic or semiautomatic means of large quantities of data in order to discover meaningful patterns and rules.

Data Miners are the people who apply this potent mixture of massive computing power, clever algorithms, business knowledge and human intuition.

What Data Mining Can Do ?

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 144

Activities that are to extract meaningful new information from the data.

The activities are :1. Classification2. Estimation3. Predication4. Affinity grouping or association rules.5. Clustering 6. Description and visualization

Classification

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 155

e.g. i. Classifying credit applicants as low,

medium or high risk.ii. Assigning customers to predefined

customer segments.iii. Assigning key-words to articles as they

come off the news-wire.

Estimation/Prediction

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 166

Use modeling to estimate the probability

e.g. estimating the value of a real estate.(MLS in US)

Prediction: It can be thought as combination of classification and estimation but the difference is in emphasis i.e. we wait to see if our prediction is correct.

E.g. Predicting if customer will move away

Affinity Grouping or Association Rules

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 177

To determine what things go together.e.g. Retail Chains can arrange items together

that are purchased together.Used to cross sell different opportunities.

Clustering

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 188

Task of segmenting a diverse group into a number of more similar subgroups or clusters

Different from Classification is absence of predefined classes rather records are grouped together on the basis of self-similarity

Normally a first step before doing something else e.g. Before ‘What type of promotion works best for customer ?’ divide people into clusters of similar buying habits.

Data Mining Discovering Knowledge based on Data

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 199

Descriptive ModelsDiscovery of patterns & relationships in the underlying datae.g. A customer who purchases diapers is 3 times more

likely to buy beer.e.g. There is a cluster of households w $60-80K incomes and

2 cars (more than ($60-$80k and 1 or 3 cars, or 2 cars w $40-60K or $80-100k) who have recently bought life insurance.

Predictive Models & Anomaly DetectionPredictions of trends & behaviors;

Noticing deviations from those predictionse.g. How much profit will this customer generate? Is this

credit card stolen? Uses

Sales & Marketing, Diagnosis, Fraud Detection, …

Data mining

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11010

Simply describe what is going on complicated database to better understanding of people products or process that produced the data.

Data visualization : one powerful form of descriptive data mining. i.e. one picture is worth many thousand association

In business context knowledge out of data-mining can be used to lower costs or raise revenue.

Data mining as Research Tool

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11111

Pharmaceutical Industry to develop leads on binding molecules.

Bioinformatics is based on data mining which interprets data being generated by high throughput screening and mapping of the entire human genetic sequence.

Data Mining for Marketing

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11212

Used to do target marketing for right customers.

Prospective targets of a marketing campaign.

Identify trends/events that correlate

Data Mining For CRM

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11313

To make the customer data into meaningful information

Good CRM means providing a good image of company across multiple channels.

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11414

BI Definition

Business intelligence (BI) is a broad category of applications and technologies for gathering, storing, analyzing, and providing access to data to help enterprise users make better business decisions. BI applications include the activities of decision support systems, query and reporting, online analytical processing (OLAP), statistical analysis, forecasting, and data mining.

BI Applications

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11515

Business intelligence applications can be: Mission-critical and integral to an

enterprise's operations or occasional to meet a special requirement

Enterprise-wide or local to one division, department, or project

Centrally initiated or driven by user demand

BI Goal

04/09/2304/09/23BI - I Lecture 1BI - I Lecture 11616

Main business intelligence goal is to provide sufficient information for making business decisions. Depending on the aim of the business decision, business intelligence methods can provide information about company customers, market trends, effectiveness of marketing campaigns, companies competitors, or even predict future activities.

Recommended