(, stascal disclosure control) - (Risk) and (Ulity) (Synthec Data) 4. 5.

Size: px

Start display at page:

Download "(, sta*s*cal disclosure control) - (Risk) and (U*lity) (Synthe*c Data) 4. 5."

웅 사
6 years ago
Views:

1 1 (, ), ( )

2 (, sta*s*cal disclosure control) - (Risk) and (U*lity) (Synthe*c Data) 4. 5.

3 3 1.

4 ,. 3. K

5 + [ ] 5 ' ', " ", " ". (SNS), '. K KT,, KG (PG), 'CSS'(Credit Scoring System)....,,,.

6 + 6? 1. ( ): 2. ( ) ( ). ( 2 2 ) 2. : 1. " ", ( ).( 2 1 ) 3. : 6. " " ( ). ( 2 6 )

7 + k- 7 ( ) [ ]

8 + 8

9 9 / / :. (2013) 2,.

10 10 ( DB)/ KCB- : DB

11 3 KCB 1 ( 31 ) KCB KCB DB * DB ( ) ( ) [ : ] 11

12 12 / - KCB DB/.

13 3. K- 13

14 + 1: (disclosure Risk) 14 k-, l-, t- ( )

15 + k-, l-, t- : 15 [ ( ) [ ],.]

16 + : k- l- 16 Key 1 Key 2 k-, l

17 + (general) 17 : : (intruder) k- : =

18 + 2: 18

19 1. (recording):. 5. / (top- down coding) (top coding) 2. (local suppression): (micro- aggrega*on): k.. 19

20 + ( ) k- 20 K-.. On k-anonymity and the Curse of Dimensionality VLDB 2005, By Charu C. Aggarwal [G-citation number 572]

21 + 1: k,, 21 : German Credit Data : 20, : 1000 : Rondom Forest X1 : account balance (4 levels), X2 : credit history (5 levels) X3 : Purpose (10 levels), X4 : Savings account (5 levels) R sdcmicro local suppression k - ( ) recording k-

22 + k=2, key =X1~X5: 5 k- obs.no X1 X2 X3 X4 1 <0DM critical account NA no savings account 2 0<=...<200DM credits paid back till now radio/television <100DM 3 no account critical account NA <100DM 4 <0DM credits paid back till now furniture/equipment <100DM 5 <0DM delay in paying off NA <100DM 6 no account credits paid back till now NA no savings account 7 no account credits paid back till now NA 500<=...<1000 DM 8 0<=...<200DM credits paid back till now used car <100DM 9 no account credits paid back till now NA >=1000DM 10 0<=...<200DM critical account new car <100DM 11 0<=...<200DM credits paid back till now new car <100DM 12 <0DM credits paid back till now business <100DM 13 0<=...<200DM credits paid back till now radio/television <100DM 14 <0DM critical account new car <100DM 15 <0DM credits paid back till now new car <100DM 16 <0DM credits paid back till now NA 100<=...<500DM 17 no account critical account radio/television no savings account 18 <0DM no credit taken business NA 19 0<=...<200DM credits paid back till now used car <100DM 20 no account credits paid back till now radio/television 500<=...<1000 DM 22

23 + k=5, key =X1~X20: 20 k- 23 obs.no X1 X2 X3 X4 1 <0DM critical account radio/television no savings account 2 0<=...<200DM credits paid back till now NA <100DM 3 NA critical account NA <100DM 4 <0DM credits paid back till now furniture/equipment <100DM 5 <0DM NA NA <100DM 6 no account credits paid back till now NA no savings account 7 no account credits paid back till now NA 500<=...<1000 DM 8 NA credits paid back till now used car <100DM 9 no account credits paid back till now radio/television NA 10 NA critical account NA <100DM 11 0<=...<200DM credits paid back till now new car NA 12 <0DM credits paid back till now NA <100DM 13 0<=...<200DM credits paid back till now radio/television <100DM 14 <0DM critical account NA NA 15 <0DM credits paid back till now new car <100DM 16 <0DM credits paid back till now NA NA 17 no account critical account radio/television no savings account 18 NA NA NA NA 19 0<=...<200DM credits paid back till now NA <100DM 20 no account credits paid back till now radio/television 500<=...<1000 DM

24 + Summary on missing rates: k- 24 Key K X1 X2 (%) X1-X5 k= k= k= X1-X10 k= k= k= X1-X15 k= k= k= X1-X20 k= k= k=

25 + 2: k-, (microaggregation) 25 R sdcmicro microaggregation : German credit data (3 ) simulated data - X1, X2, X3 : German credit data (X1 : duration, X2: credit amount, X3 : Age) - X4, X5, X20 [1,10] uniform : 1000

26 k=10, key = X1~X20 obs. No Du CA Age , , , , , , , , , , , , , , , , , , , , obs. No Du CA Age , , , , , , , , , , , , , , , , , , , ,

27 10- obs.no Du CA Age , , , , , , , , , , obs.no Du CA Age , , , , , , , , , ,

28 ( ) k=10 169, 726, 916

30 30 4. (synthetic data)

31 31 Synthetic Data (synthe*c data, ), - (par*ally synthe*c data) - (fully synthe*c data)

32 (synthe*c): =, =.[ ] 32

33 -, key- synthe*ze - 33

34 34 Synthetic Data : 1) (re- iden*fica*on).. 2) [ ] 3).

35 :. (. ) -., Synthe*c data. 35

36 36 Synthetic data SBB(SIPP Synthe*c Beta): Federal Privacy Council Census of Bureau Census of Bureau Survey of Income and Program Par8cipa8on(SIPP) Social Security Administra8on(SSA)/Internal Revenue Service(IRS)

37 : hxps:// surveys/sipp/ 3. Census of Bureau, / 37

38 Synthe*c SIPP data 1. Applica*on form,, SIPP (123 ) account SAS Stata SSB data 4. (SAS Stata code) 38

39 39 Synthetic Data : German Credit Data Data : German credit data( =1000) 900 : training data 100 : test data Synthetic variable y (credit status : good, bad) 3 (duration, credit amount, age) Models for synthesis : f(y ): logistic regression : f(c1 ), f(c2 ), f(c3 ): linear regression

40 Synthe*c data training synthesis 2. training synthetic data set 3. Synthetic data set y (logistic regression) 4. 3 model original training data Testing (100 ) 40

41 Synthe*c data : Original Training Data obs.no y duration credit.amount age account balance Credit history purpose 1 good 6 1, <0 DM critical account radio/television 2 bad 48 5, <=...<200 DM credits paid back till now radio/television 3 good 12 2, no account critical account education 4 good 42 7, <0 DM credits paid back till now furniture/equipment 6 good 36 9, no account credits paid back till now education 7 good 24 2, no account credits paid back till now furniture/equipment 8 good 36 6, <=...<200 DM credits paid back till now used car 9 good 12 3, no account credits paid back till now radio/television 10 bad 30 5, <=...<200 DM critical account new car 41

42 Synthe*c Data obs.n o 1 y goo d duratio n credit.amou age nt account balance Credit history purpose 33 2, <0 DM critical account radio/television 2 bad 32 4, goo d goo d goo d goo d goo d goo d 0<=...<200 DM credits paid back till now radio/television no account critical account education 41 4, <0 DM 24 8, no account 9 2, no account 29 5, <=...<200 DM 32 3, no account 10 bad 19 3, <=...<200 DM credits paid back till furniture/ now equipment credits paid back till education now credits paid back till furniture/ now equipment credits paid back till used car now credits paid back till radio/television now critical account new car 42

43 Synthe*c data Synthetic 100 set, ( ) (confusion matrix) original data synthetic data model true true true bad true bad good good pred. bad (0.418) (0.497) pred. good (0.418) (0.497) model original data synthetic data

44 44 K- Synthetic Data K

45 45 5. (Differential Privacy)

46 + (Differential Privacy) : 라플라스기계 (Laplance machine) 3. Local differential privacy 4. 5.

47 + 47 : 1. ( ) (Q) 2. R

48 + 48

49 A,, ( ). : NOISE NOISE. 49

50 + 1: 50

51 : 2: : 51

52 52

53 Post- processing Invariance: ε-dp Algorithm Algorithm Output ε- DP [ K. Chaudhuri A.D. Sarwate ] 53

54 Composi*on: A 1 (, ) ε 1 - DP A 2 (, ) ε 2 - DP A 1 (D) A 2 (D) [ K. Chaudhuri A.D. Sarwate ] 54

55 Dependency to :.,, (classifier),. 55

56 + Local Differential Privacy 56 DP( ) central. Epsilon noise. Local DP central noise central.

57 + 3: randomized response 57 Randomized response: : / O/X. O. Google Chrome : /.

58 + 4: NYC taxicab data set 58 Riding with the stars, from research.neustar.biz

59 NY taxicab dataset : pickup and drop off times, locations, fare and tip amounts de-anonymize, (Driver privacy) privacy (Passenger privacy in the NYC taxicab dataset). : the picture, some information from celebrity gossip blogs

60 + 60 Jessica Alba,, ($9), ($0)

61 + Differentially Privatized Trip (drop-off) 61 Total Fare: $25 - $30 Tip Amount: $6 - $8

62 + NYC taxicab data: local DP 62

63 + K- 63 Differentially Privatized Synthetic Data: 1. DB Local DP ε-ldp /

64 2. LDP. / / ε-ldp / ε-ldp / ε-ldp / 3. DP/local DP procedure 64

65 65 6.

66 / 3... K- 4.,

67 67-5. / - /

68 Thank you 68

사회통계포럼

사회통계포럼 wcjang@snu.ac.kr Acknowledgements Dr. Roger Peng Coursera course. https://github.com/rdpeng/courses Creative Commons by Attribution /. 10 : SNS (twitter, facebook), (functional data) : (, ),, /Data Science