Top Banner
TM May 2013 datastereo.com 1 5 SLIDES TO INTRODUCE DATA MINING
19

15 Slides to Introduce Data Mining

May 24, 2015

Download

Technology

Yves Sukhu

There is a lot of talk these days about "Big Data" and data mining. I put this non-technical presentation together to introduce data mining to business users who are looking to separate the hype from the reality. It discusses what makes something data mining versus traditional information retrieval or business intelligence.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: 15 Slides to Introduce Data Mining

TM

May 2013 datastereo.com

15 SLIDES TO INTRODUCE

DATA MINING

Page 2: 15 Slides to Introduce Data Mining

First of all…is data mining new?

datastereo.comMay 2013

Page 3: 15 Slides to Introduce Data Mining

TM

History“Data mining” has been around for a long time

May 2013

• Early civilizations studied the weather to figure out the best time(s) of year to plant crops.

• Speaking of the weather, forecasters over the centuries have attempted to predict the weather by studying years of recorded weather data.

• Economists have collected and analyzed social and economic data to make financial forecasts .

datastereo.com

Page 4: 15 Slides to Introduce Data Mining

TM

So What’s All This “Big Data” Hype?There is a lot of data out there and it’s growing

May 2013

• The human race is creating data at an increasingly faster pace.

• It is estimated that the amount of data stored in all the world’s databases doubles every 20 months.

• Every day, we create 2.5 quintillion bytes of data.– IBM® Corporation.

datastereo.com

Page 5: 15 Slides to Introduce Data Mining

2,500,000,000,000,000,000That’s

bytes every single day.

datastereo.comMay 2013

Page 6: 15 Slides to Introduce Data Mining

TM

That’s a Lot of DataWhere is it all coming from?• A lot of it is coming from you:

– GPS signals from your cell phone.– Photos uploaded to Facebook.– Posts to Twitter.– Purchase transactions when you buy online.

• Other sources include sensors, satellites, network devices, and basically anything that captures digital information.

May 2013 datastereo.com

Page 7: 15 Slides to Introduce Data Mining

TM

How Do We Make Sense of It All?Data mining to the rescue.• Data mining allows us to ask questions about all

this data, and seek out answers.– What is the probability that a category 5 hurricane

will strike the northeast U.S. in the next year, and cause $10 billion or more in damage?

– Which of my customers are most likely to respond to a new sale promotion?

May 2013 datastereo.com

Page 8: 15 Slides to Introduce Data Mining

How is data mining different fromtraditionalinformation retrieval?

May 2013 datastereo.com

Page 9: 15 Slides to Introduce Data Mining

TM

It’s All About KnowledgeLooking for data vs. looking for knowledge.

May 2013

Examples of

Traditional Information Retrieval

(looking for data)

• Total sales for the last 12 months.

• Show a list of stocks whose price increased in the last 30 days.

• Number of employees in each department.

Examples of

Data Mining

(looking for knowledge)

• What are predicted sales for the next 12 months?

• What stocks are most likely to increase by 10% in the next 30 days?

• Which employees are most likely to resign?

datastereo.com

Page 10: 15 Slides to Introduce Data Mining

TM

How About a DefinitionLike we said, it’s all about knowledge.

May 2013

Data mining is the automated (or semi-automated) process of discovering and

extracting knowledge from data.

datastereo.com

Page 11: 15 Slides to Introduce Data Mining

TM

3 Things That Make It Data MiningAs opposed to just retrieving data.

1. You are trying to discover knowledge from your data, not just retrieving or summarizing data.

2. The process of discovery is automated or semi-automated through any number of data mining computer tools and techniques.

3. There is (usually) some competitive, economic, or social value in the knowledge being sought.

May 2013 @DataStereo

Page 12: 15 Slides to Introduce Data Mining

How is data mining different fromtraditionalbusiness intelligence?

May 2013 datastereo.com

Page 13: 15 Slides to Introduce Data Mining

TM

It Depends Who You AskSummarization vs. discovery.• Some would say data mining is part of BI; others

would say data mining is a separate discipline.• It’s probably fair to say that they do overlap.• There is no perfect definition, but we like to think of

it in terms of summarization vs. knowledge.– BI tends to be about summarizing data and data mining

is about discovering knowledge.

May 2013 datastereo.com

Page 14: 15 Slides to Introduce Data Mining

TM

I’ve Heard Some Other Terms TooData mining is an overused term.• Decision support, data warehousing, data analytics,

machine learning are often used synonymously with data mining.

• The term tends to get associated with any kind of computer processing that involves really large amounts of data.

• Just remember, data mining is about discovering knowledge.

May 2013 datastereo.com

Page 15: 15 Slides to Introduce Data Mining

Who is usingdata mining?

datastereo.comMay 2013

Page 16: 15 Slides to Introduce Data Mining

TM

Data Mining Is EverywhereIn all industries and all roles.

May 2013

Industries

• Retail• Telecommunications• Pharmaceuticals• Financial services• Hospitality and travel• Media and entertainment• Consumer products

Job Roles

• Marketing• Sales• Risk management• Human resources• Manufacturing• Logistics• Customer service

datastereo.com

Page 17: 15 Slides to Introduce Data Mining

Sounds interesting?How do I learn more

datastereo.comMay 2013

about data mining?

Page 18: 15 Slides to Introduce Data Mining

Stay tuned!More to come…

May 2013 datastereo.com

Page 19: 15 Slides to Introduce Data Mining

TM

May 2013 datastereo.com

We’re buildingan app for you.Learn data mining fast, easily, and socially.

Sign-up for the beta at:www.datastereo.com