Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Seminar - Data Mining

No description

Ka Howe Yapp

on 1 October 2012

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Seminar - Data Mining

Seminar Presentation
By Yapp Ka Howe Application of
Data Mining in
Retail Business to
Facilitate Marketing
Strategy. Why Data Mining? What is Data Mining? Methods of Data Mining Phase 1 : Selection Application of Data Mining to Marketing Data Mining Process Application of Data Mining
to Final Year Project I have data,
but I know nothing. Discover fact, pattern Classification Five Phases Select the required data
based on criteria Communicate with customers Fast Food Restaurant System > > Raw Data Data Mining Knowledge Clustering Association Sequential Pattern Classification Two steps: learning > classification Can be used to predict such as: Association > 70% Confidence Discover association rules:
If A, then B with x% confidence. Clustering Segments related data based on similarity. Sequential Pattern Applicable when customer identity is known. Customer bought A also bought B later. In separate transactions. Marketing strategy planning? > Customers data
> Transactions data 5 Phases Phase 2 : Pre-processing Remove noise such as data inconsistency
and unnecessary data. Sample patient data. Phase 3 : Transformation Generate analytical variables Date of birth --> Age Normalization Rescale values, σ = 1 and μ = 0 Avoid inclination towards variable with greater range Phase 4 : Data Mining Preferably starts with something simple Classification > clustering > association > others Iterative process : Zoom in Zoom out Phase 5 : Evaluation a.k.a interpretation Discovered patterns are explained & described Evaluate result quality Iterate the process if needed Prospecting with
Data Mining Identify potential customers Existing customer characteristics:
- 61% college educated
- 47% have professional jobs
- 23% earns > RM75,000 annually
- 9% earns > RM100,000 annually Predicting Response Rate
with Data Mining Budget: RM 30,000 Direct Marketing campaign Cost per contact: RM1 Gross profit per response: RM 45 Contacting top 30,000 prospects Average response rate: 2% Cost: RM 30,000
Responders: 600
Gross Profits: RM 27,000
Loss: RM3,000 Contacting top 10,000 prospects Cost: RM 10,000
Responders: 300
Gross Profits: RM 13,500
Revenue: RM3,500 Average response rate: 3% Differential Response Analysis
with Data Mining Prospects with highest score may become customers even without contacting them. Benefits & Drawbacks
of Data Mining What is Data Mining Good About? Customer segmentation, 20/80 principle. Sales prediction helps new location targeting. Get closer to customers. Others: Counter terrorist, military, estimate loan risk What Challenges Data Mining Pose? Customer information trading - harm privacy. Lack of security to protect information. Used for unethical purposes. Reporting Module Customer Buying Pattern Report Customer Spending Behavior Report Order Performance Report Sample values {4, 78, 1205, 0.034, 0.23, 677}
Full transcript