Practical Applications of Homomorphic Encryption

Practical Applications of Homomorphic Encryption

Practical Applications of Homomorphic Encryption

Birmingham, Paul, News Anchor has reference to this Academic Journal, PHwiki organized this Journal Practical Applications of Homomorphic Encryption Kristin LauterCryptography Research GroupMicrosoft ResearchCrypto Day 2015May 13, 2015Protecting Data via Encryption: Homomorphic encryption Put your gold in a locked box. Keep the key. Let your jeweler work on it through a glove box. Unlock the box when the jeweler is done!Homomorphic Encryption: additioncomputecomputeencryptencrypta, ba+ bE(a), E(b)

Grambling State University US

This Particular University is Related to this Particular Journal

Homomorphic Encryption: multiplicationcomputecomputeencryptencrypta, ba x bE(a), E(b)Operating on encrypted data“Doubly” homomorphic encryptionAmerican Scientist, Sept/Oct 2012 Secure Genome Analysis CompetitioniDASH Privacy & Security Workshop 2015 Sponsored by NIH (National Institutes of Health)Submission deadline: Feb 28 2015Workshop: March 16, 2015 UCSD Medical Education in addition to Telemedicine BuildingMedia coverage in GenomeWeb, Donga Science, NatureTeams from: Microsoft, IBM, Stan as long as d/MIT, UCI, University of TsukubaTwo Tracks: MPC in addition to HEChallenges: GWAS in addition to Sequence Alignment

Donga Science, March 13, 2015 MS DNA (28) . 1 (MS) . MS DNA ‘ (Secure Genome Analysis Competition)’ . ., Nature,

Why the excitementFundamental Problem: privacy protection Burgeoning genome sequencing capabilityExplosion of scientific research possibleHigh risk as long as personal privacy Fundamental Progress through interactionComputer ScientistsMathematiciansBioin as long as maticiansPolicy-makersGenomic RevolutionFast drop in the cost of genome-sequencing2000: $3 billionMar. 2014: $1,000Genotyping 1M variations: below $200Unleashing the potential of the technologyHealthcare: e.g., disease risk detection, personalized medicine Biomedical research: e.g., geno-phono associationLegal in addition to as long as ensicDTC: e.g., ancestry test, paternity test Genome PrivacyPrivacy risksGenetic disease disclosureCollateral damageGenetic discriminationGr in addition to Challenges:How to share genomic data or learning in a way that preserves the privacy of the data donors, without undermining the utility of the data or impeding its convenient disseminationHow to per as long as m a LARGE-SCALE, PRIVACY-PRESERVING analysis on genomic data, in an untrusted cloud environment or across multiple users

Data access in addition to sharing requirementsAllow access to researchers to large data setsSecure Genome Wide Association Studies (GWAS)Desire as long as centrally hosted, curated dataProvide services based on genomic science discoveriesTwo scenarios as long as interactions:Single data owner (one patient, one hospital)Multiple data owners (mutually distrusting)Two Challenges!Challenge 1: Homomorphic encryption (HE) based secure genomic data analysisTask 1: Secure Outsourcing GWASTask 2: Secure comparison between genomic dataChallenge 2: Secure multiparty computing (MPC) based secure genomic data analysis (two institutions) Task 1: Secure distributed GWASTask 2: Secure comparison between genomic dataData Source 200 Cases from Personal Genome Project (PGP)PGP: launched by Harvard Medical School 200 Controls were simulated based on the haplotypes of 174 individuals from population of International HapMap Project ( individual genomes (hu604D39 with 4,542,542 variations in addition to hu661AD0 with 4,368,847 variations comparing to the reference human genome) were r in addition to omly selected from PGP

Results as long as Task 1.1: Minor Allele Frequency (training dataset with 311 SNPs, time in seconds)Results as long as Task 1.2 (Hamming)Results as long as Task 1.2 (Approximate Edit distances)An approximate algorithm (with about 22% error), which was not considered in the competition.

WinnersTask 1.1: Stan as long as d/MITTask 1.2: Hamming distance: IBMTask 1.2: Approximate Edit distance: MicrosoftPractical problems:Cleaning/curating dataEncoding dataTrade-offs in computation time vs. memoryParameter selection: challenging to optimize in addition to automateFollow-upReport to NIHSpecial Issue in Biomedical In as long as matics in addition to Medical Decision-makingPapers from each team describing their submissions

What scenarios make sense as long as HEPrivate, personalized cloud servicesIdeally combined with storage, or asynchronized accessMultiple parties upload data, only designated parties access resultsLong-term storage is desirableCryptographic Cloud ServicesHosted enterprise scenarios as long as storage in addition to computationScenarios: Private cloud servicesDirect-to-patient servicesPersonalized medicineDNA sequence analysisDisease prediction Hosted databases as long as enterpriseHospitals, clinics, companiesAllows as long as third party interactionOutsourcing computation

Birmingham, Paul Jim Parisi Show  - KNST-AM, The News Anchor

Demo: Will you have a heart attackOnline service running in Windows AzurePatient enters personal info on local machine: weight, age, height, blood pressure, body mass indexData is encrypted on local machineEncrypted data is sent to the cloudValue of prediction function is computed on encrypted dataEncrypted result is sent back to the patientPatient enters key to decrypt answer.Evaluation takes 0.2 seconds in the cloud!All data uploaded to the server encrypted under Alice’s public or private keyCloud operates on encrypted data in addition to returns encrypted predictive resultsProcessing of encrypted medical dataHealth monitorLab resultsHealth monitor

Scenario as long as genomic dataTrusted partyhosts data in addition to regulates accessUntrusted cloud service Stores, computes on encrypted dataResearcher: requests encrypted results of specific computationsRequests as long as decryption of results (requires a policy)Homomorphic Encryption from RLWEUses polynomial rings as plaintext in addition to ciphertext spacesWhat kinds of computationBuilding predictive modelsPredictive analysisClassification tasksDisease predictionSequence matchingData quality testingBasic statistical functionsStatistical computations on genomic data

What are the Costs Challenges ObstaclesFor homomorphic encryptionStorage costs (large ciphertexts)New hard problems (introduced 2010-2015)Efficiency at scale (large amounts of data, deep circuits)For Garbled CircuitsHigh interaction costsB in addition to width useIntegrate with storage solutionsChallenges as long as the future:Public Databases: multiple patients under different keysMore efficient encryption at scaleIntegrate with other crypto solutionsExp in addition to functionalityAttack underlying hard problemsJoint work with: in addition to thanks to iDASH in addition to co-authors as long as selected slides Can Homomorphic Encryption be PracticalKristin Lauter, Michael Naehrig, Vinod Vaikuntanathan, CCSW 2011ML Confidential: Machine Learning on Encrypted Data Thore Graepel, Kristin Lauter, Michael Naehrig, ICISC 2012Predictive Analysis on Encrypted Medical DataJoppe W. Bos, Kristin Lauter, in addition to Michael Naehrig, Journal of Biomedical In as long as matics, 2014.Private Computation on Encrypted Genomic DataKristin Lauter, Adriana Lopez-Alt, Michael Naehrig, GenoPri2014, LatinCrypt2014.Homomorphic Computation of Edit DistanceJung Hee Cheon, Miran Kim, Kristin Lauter, WAHC, FC 2015

Birmingham, Paul News Anchor

Birmingham, Paul is from United States and they belong to Jim Parisi Show – KNST-AM, The and they are from  Tucson, United States got related to this Particular Journal. and Birmingham, Paul deal with the subjects like Federal Government and Politics; National News

Journal Ratings by Grambling State University

This Particular Journal got reviewed and rated by Grambling State University and short form of this particular Institution is US and gave this Journal an Excellent Rating.