PRIVACY PRESERVING BACK-PROPOGATION NEURAL NETWORK LEARNING MADE PRACTICAL WITH CLOUD COMPUTING

Preview:

DESCRIPTION

 

Citation preview

WELCOME

PRESENTED BY,THUSHARA.M

M.Tech CSISROLL NO:18

PRIVACY PRESERVING BACK-PROPOGATION NEURAL NETWORK LEARNING IN CLOUD

COMPUTING

3

IntroductionLiterature reviewContributionsModels and assumptionsTechnique preliminariesProposed schemePerformance evaluationConclusionReferences

CONTENTS

4

Neural networks.Back-propogation.Improves the accuracy.Joint/Collaborative learning.

INTRODUCTION

5

Challenges: To protect each participant’s private data set and

intermediate results.

The computation/ communication cost introduced to each participant shall be affordable.

For collaborative training, training data is arbitrarily partitioned.

INTRODUCTION(Contd..)

6

Provides privacy preservation for multiparty .

Collaborative BPN network learning over arbitrarily partitioned data.

Guarantees privacy and efficiency.

Support multiparty secure scalar product.

Allow decryption of arbitrary large messages.

CONTRIBUTONS

7

System Model: Trusted authority. The participating parties ( data owner). The cloud servers ( or cloud).

Security Model:

MODELS AND ASSUMPTIONS

8

Arbitrarily Partitioned Data Z parties (Z > 2 ) : Ps , 1 ≤ s ≤ Z. Database D with N rows : {DB1,DB2, ….. DBN}. Each row DBv ,1 ≤ v ≤ N has m attributes {xv

1 , xv2 , xv

3 …..

xvm}.

DBv = DBv1 U DBv

2 U DBv3 U ….. U DBv

z .

Each DBv, Ps has tsv attributes.

TECHNIQUE PRELIMINARIES

9

BACK –PROPOGATION NEURAL NETWORK LEARNING

TECHNIQUE PRELIMINARIES(Contd..)

10

BGN Homomorphic Encryption Operations on plaintexts to be performed on their

respective cipher texts. Public-key “doubly homomorphic” encryption

scheme(called “BGN” for short). One multiplication and unlimited number of additions. Given ciphertexts C(m1) , C(m2) and C(m^1), C(m^2 ), one

can compute C(m1 m^1 + m2m^2) without knowing the plaintext.

TECHNIQUE PRELIMINARIES(Contd..)

11

PROBLEM STATEMENT 3 layer (a-b-c configuration) neural network . N samples for learning data set . Arbitrary partitioned into Z( Z≥2) subsets.

SCHEME OVERVIEW Each party encrypt her/his input data set. Participants upload the encrypted data to cloud. Cloud servers perform the operations. Secret sharing algorithm.

PROPOSED SCHEME

12

PRIVACY PRESERVING MULTIPARTY NEURAL NETWORK LEARNING

PROPOSED SCHEME(Contd..)

13

PROPOSED SCHEME(Contd..)

14

SECURE SCALAR PRODUCTION AND ADDITION WITH CLOUD

Algorithm 3: Secure Scalar Product and Addition Key Generation. Encryption. Secure Scalar Product. Secure Addition. Decryption.

PROPOSED SCHEME(Contd..)

15

SECURE SHARING OF SCALAR PRODUCT AND SUM

PROPOSED SCHEME(Contd..)

16

APPROXIMATION OF ACTIVATION FUNCTION

PROPOSED SCHEME(Contd..)

17

Experimental Evaluation Experiment Setup• Amazon EC2 cloud.• 10 nodes with 8-core 2.93-GHz Intel Xeon CPU.• 8-GB memory.• Testing data sets(Iris,kr-vs-kp,diabetes).

PERFORMANCE EVALUATION

18

EXPERIMENTAL RESULT

PERFORMANCE EVALUATION(Contd..)

19

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

20

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

21

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

22

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

23

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

24

EXPERIMENTAL RESULT (Contd..)

PERFORMANCE EVALUATION(Contd..)

25

ACCURACY ANALYSIS Accuracy loss in approximation of activation function. Maclaurin series used – accuracy can be adjusted by

modifying number of series terms.

PERFORMANCE EVALUATION(Contd..)

26

Secure and practical multiparty BPN network learning.

Cost independent of number of parties.

Scalable efficient and secure.

CONCLUSION

27

1) N. Schlitter A Protocol for Privacy Preserving Neural Network Learning on Horizontal Partitioned Data, Proc. Privacy Statistics in Databases (PSD ’08), Sept. 20082) T. Chen and S. Zhong, Privacy-Preserving Backpropagation Neural Network Learning,IEEE Trans. Neural Network, vol. 20, no. 10, Oct. 2000,pp. 1554-15643) A. Bansal, T. Chen, and S. Zhong, Privacy Preserving Back-Propagation Neural Network Learning over Arbitrarily Parti-tioned Data,Neural Computing Applications,vol. 20, no. 1, Feb. 2011, pp. 143-150, 4) D. Boneh, E.-J. Goh, and K. Nissim, Evaluating 2-DNF Formulas on Ciphertexts,Proc. Second Int’l Conf. Theory of Cryptography (TCC ’05), pp. 325-341, 2005.

REFERENCES

THANK YOU

Recommended