Closed

# Java Project to implement Data Mining Concept (Perturbation)

This project was awarded to Cubeheap3 for \$55 USD.

\$30 - \$60 USD
5
###### Project Description

Data Perturbation is a technique where noise is added to data to make it impossible to derive accurately the values of sensitive/confidential data.
The additive noise still permits the aggregate information to be read, about the overall collection of data, but does not give away accurate values.
The noise is a small randomly generated (or using certain algorithms), and added to the data.
Hence, by this method we protect individual info and releases aggregate info at same time.

Algorithm:

1. Consider a data base D with n tuples t = {t1, t2....tn}.
Each tuples contains Set of attribute A = {A1, A2.....Am} A € ti.

2. Find the sensitive attribute SAR for all SAR € Ai € A (i=1,2...m).

3. Calculate Average for all SAR (i).

4. Do

Initialize the value i = (1, 2....n).
Check if (Average ≥ SAR (i.) to count the all values C1. Calculate M1 = (2*Average)/C1
Replace SAR (i). With M1.
Check if (Average ≤ SAR (i.) to count the all values C2. Calculate M2 = (2*Average)/C2
Replace SAR (i). With M2.
Increment the value of i.

5. While (i ≥ n) 6. End.

## Hire Freelancers who also bid on this project

• Forbes
• The New York Times
• Time
• Wall Street Journal
• Times Online