Frequent Itemset Genration

Market Basket Analysis

A solution to your marketing analysis problems

Real business scenarios and

problemsWhy didn’t Ithought of this beforeIs it possible,

with the number of products we deal in

What can you do?

• Experiential Approach: Use your experiential learning

• Algorithms Approach: Try a new algorithm from one’s offered by geek squad

Experiential Approach• Product

decision• Managemen

t• Marketing• Employees• Salary• Handling

Competition

Too much to handle??

Algorithms Approach

Sounds even more

Troublesome??

Try Experts

• Pioneers in Marketing and Retail chain solutions

• We value our customers

• We believe in excellence

• Customized solution for problems

• We value your business like our brand

Last few Achievements

Partnered with 50 retails firms which are now enjoying status of being in fortune 500

Signed a deal with the largest retail chain giant “Falmart” for their global chain

Rated one of the best firms in Professional Retail Magazine for retail chain solutions

Used with client's permission

Market Basket Analysis

CB A

Problem at hand: Our distinct products:Product AProduct BProduct C

Output:Customer who buy A also buy B.

Solution:Try bundling A&BPlace the two on adjacent shelvesDiscounts on the package deal

A&B

Benefits of Implementation

All the products that are sold together Promotions Bundling offers Better inventory planning

ResultsHappy customer

Decent ratings

Long term relationship

Boost in sales Term

1Te

rm2Te

rm3Te

rm4

00.5

11.5

22.5

33.5

44.5

Profits

Profits

More Results

Customer Satisfaction

* Offers* Purchase Satisfactions

Better Inventory Management

* Low shelve time* Better planning

Boost in sales

* Product line expansion opportunity* Investment opportunity* …

• Profits• Custome

r base expansion

• Efficiency

Frequent Item-Sets item_1,

item_9, item_35, item_39, item_42

item_2, item_7, item_29

item_3, item_5, item_22

Thank You!

Appendix Association Rules

Apriori Algorithm Eclat FP-Tree

Apriori Algorithm for frequent item sets mining and

association rule learning over database of transactions.

Generate k level of all possible keys and then eliminate the k-1 level combinations from the rejected list for the k-1 level.

Perform the same recursively, to get the final output.

FP-Tree Efficient algorithm; Uses only two database

scans FP-Tree data structure Efficient implementation using hash tree