privacy-preserving self- organizing map shuguo han and wee keong ng center for advanced information...

16
Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological University, Singapore (DaWak 2007) 1 2009/11/02

Upload: sharyl-fox

Post on 18-Jan-2016

214 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-Preserving Self-Organizing Map

Shuguo Han and Wee Keong Ng

Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

University, Singapore(DaWak 2007)

12009/11/02

Page 2: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Outline

Introduction SOM Privacy-preserving SOM protocol Conclusion

22009/11/02

Page 3: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Introduction various data mining algorithms have been

enhanced with a privacy preserving version for horizontally and/or vertically partitioned data

propose a protocol for privacy-preserving self-organizing map for vertically partitioned data involving two parties.

2009/11/02 3

Page 4: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

SOM Self-organizing map (SOM) is awidely used algorithmfor

transforming data sets to a lower dimensional space to facilitate visualization

To projection of the data set while preserving the topological properties of the data set.

SOM is a feed-forward neural network without any hidden layer adjusting input.

2009/11/02 4

Page 5: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

SOM

Competition phase, 不斷學習使輸出與目標值能達到相同值後結束

input data X(t) = [Xi(t),X2(t), . . . , Xd(t)] each neuron’s weight vector

(randomly initial weight ) Wj(t) = [Wj,1(t),Wj,2(t), . . . ,Wj,d(t) ]

2009/11/02 5

Page 6: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

SOM

Euclidean distance:

Winner neuron:

Update weight vector

2009/11/02 6

Z為學習函式 ,越大表學習越快 ,一般介於 0~1之間

Page 7: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Protocol 1. Privacy-Preserving Self-Organizing Map

, the weight vector holds two private component vector. At step t=0 where and are securely and randomly generated respectively

2009/11/02 7

Page 8: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

input data X = (X1,X2, . . . , Xd) from feature space

The different between SOM and stand SOM is that the subprotocol are required to perform some computations securely

2009/11/02 8

Page 9: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Protocol 2. Secure Computation of Closest Cluster Protocol

=> to find winner neuron by

applying the secure scalar product protocol [2,4].

2009/11/02 9

Page 10: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

2009/11/02 10

Page 11: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Correctness

2009/11/02 11

Page 12: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Protocol 3. Secure Weight Vector Update Protocol

2009/11/02 12

// adjust all neuron’s weight vector

// j is how many neurons in this grid// i is how many attributes of input

Page 13: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Correctness

2009/11/02 13

Page 14: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Protocol 4. Secure Detection of Termination Protocol

2009/11/02 14

Page 15: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Privacy-preserving SOM protocol

Correctness

stop

2009/11/02 15

Page 16: Privacy-Preserving Self- Organizing Map Shuguo Han and Wee Keong Ng Center for Advanced Information Systems, School of Computer Engineering,Nanyang Technological

Conclusion

(1) to securely discover the winner neuron from data privately held by two parties

(2) to securely update weight vectors of neurons

(3) to securely determine the termination status of SOM.

2009/11/02 16