research-article   1160

« earlier    

[1809.09561] Evaluating stochastic seeding strategies in networks
[spooky sound] daaaannnnnngerous research!

Slightly more seriously: I've had this idea for a while, that you could use social networks to scale up re-identifying attacks a la Erlich's 2013 surnames-from-anonymous-genomes paper. Basically, you're more likely to be connected to people in a social network (of almost any kind) if you're related to them, than if they're some rando off the street. Ergo, you should be able to compute approximate distances in social networks via genetic measures (% IBD? longest IBD region? etc.) and work backwards through the network to resolve an anonymous sequence into a network node. I think the process must look something like what's described here.

I probably should think this through in more detail.

(I remain pretty confident that people like Y.E. have been talking about this with TLAs for a number of years, so this isn't really 'cutting edge' anymore. <sigh>)
dean-eckles  social-networks  research-article  arxiv 
7 days ago by arthegall
Kairouz, Oh, Viswanath, "Extremal Mechanisms for Local Differential Privacy"
"We introduce a family of extremal privatization mechanisms, which we call staircase mechanisms, and prove that it contains the optimal privatization mechanism that maximizes utility. We further show that for all information theoretic utility functions studied in this paper, maximizing utility is equivalent to solving a linear program, the outcome of which is the optimal staircase mechanism."
local-differential-privacy  differential-privacy  research-article  linear-programming  randomization 
7 days ago by arthegall
Joseph, Mao, Neel, Roth, "The Role of Interactivity in Local Differential Privacy" (arXiv)
"First, we classify locally private protocols by their compositionality, the multiplicative factor k≥1 by which the sum of a protocol's single-round privacy parameters exceeds its overall privacy guarantee. We then show how to efficiently transform any fully interactive k-compositional protocol into an equivalent sequentially interactive protocol with an O(k) blowup in sample complexity."
differential-privacy  interactivity  algorithms  research-article  arxiv  local-differential-privacy 
7 days ago by arthegall
Kairouz, Oh, Viswanath, "Differentially Private Multi-party Computation: Optimality of Non-Interactive Randomized Response"
"We study the problem of interactive function computation by multiple parties possessing a single bit each in a differential privacy setting... Our main result is the exact optimality of a simple non-interactive protocol: each party randomizes (sufficiently) and publishes its own bit. In other words, non-interactive randomized response is exactly optimal."
differential-privacy  research-article  randomization 
7 days ago by arthegall
Kairouz, Oh, Viswanath, "The Composition Theorem for Differential Privacy"
"Sequential querying of differentially private mechanisms degrades the overall privacy level. In this paper, we answer the fundamental question of characterizing the level of overall privacy degradation as a function of the number of queries and the privacy levels maintained by each privatization mechanism."
differential-privacy  research-article  privacy 
7 days ago by arthegall
Duchi, Jordan, Wainwright, "Local Privacy and Statistical Minimax Rates"
Introduces "local differential privacy," where data remains private even from the learner or data analyst.
local-differential-privacy  differential-privacy  research-article  michael-jordan  privacy 
7 days ago by arthegall
Relatedness disequilibrium regression estimates heritability without environmental bias | Nature Genetics
This paper functions well as a tutorial, on different methods for estimating "fraction of heritability explained" through different genetic methods, and how they might (or might not) be biased estimates in the presence of 'environmental' effects.
heritability  genetics  research-article  kari-stefansson 
4 weeks ago by arthegall
[1902.04114] Using Embeddings to Correct for Unobserved Confounding
"We consider causal inference in the presence of unobserved confounding. In particular, we study the case where a proxy is available for the confounder but the proxy has non-iid structure. As one example, the link structure of a social network carries information about its members."
victor-veitch  arxiv  research-article  machinelearning  confounding  causal-learning  causality 
8 weeks ago by arthegall
A Flow Procedure for the Linearization of Genome Sequence Graphs. | bioRxiv
This reminds me of a project Ted and I worked on, way back when, in Mike Stonebraker and Sam Madden's database class.
linearization  graph-genomes  bioinformatics  preprint  research-article  david-haussler  benedict-paten  genomics  sequence-analysis 
9 weeks ago by arthegall
Bulk Synchronous Parallel Computing — A Paradigm for Transportable Software | SpringerLink
I'm sure someone's thought through this already, but I wonder how well this aligns with the way we think about (e.g.) Spark
bsp  distributed-computing  research-article  parallel-computing 
9 weeks ago by arthegall
Assessing graph-based read mappers against a novel baseline approach highlights strengths and weaknesses of the current generation of methods | bioRxiv
"Graph genomes" require "graph mappers," but alignment to graph genomes doesn't have the same statistics or methods for assessing accuracy (both at the individual and global level) as traditional mapping against linear genomes.
bioinformatics  graph-algorithms  graph-genomes  sequence-analysis  alignment  research-article  preprint 
9 weeks ago by arthegall
[1512.01625] Coded MapReduce
Feels, at first glance, a lot like the 'sudoku' methods in genomics, and uses of codes in experimental designs
mapreduce  arxiv  computerscience  research-article  code  coding-theory 
november 2018 by arthegall

« earlier    

related tags

adam-klivans  adam  adaptive-algorithms  alex-smola  algebraic-geometry  algebraic-statistics  algorithm  algorithms  alignment  alzforum  analytics  annotation  anonymization  approximation  arxiv  assembly  auctions  avrim-blum  back-propagation  bayesian-methods  benedict-paten  benjamin-pierce  big-data  bioarxiv  bioinformatics  biology  black-box  book-chapter  bsp  calm  cancer  carlos-guestrin  causal-learning  causality  citations  classification  clinical  clustering  code  coding-theory  combinatorics  competition  compiler  complexity  compression  computer-algebra  computer-graphics  computer-science  computers  computerscience  concordance  confounding  congestion  costs  crowdsourcing  danielle-posthuma  data-science  data-streams  data-structures  data  databases  david-haussler  dbscan  dean-eckles  debugging  deep-learning  demographics  developer-experience  diagnostics  differential-equations  differential-privacy  diffusion  dimension  dirichlet-allocation  dirichlet-process  distributed-computing  distributed-systems  do-no-map  dynamic-problems  economics  elections  embedding  eric-jonas  erik-garrison  ethics  eugene-myers  expectation-propagation  explanation  extremism  feature-selection  formal-methods  free-resolutions  functionalprogramming  games  generalization  genetic-architecture  genetics  genomics  genotyping  gotv  gradient-descent  graph-algorithms  graph-genomes  graph-methods  graphical-models  graphics  graphs  guassian-processes  gunnar-carlsson  gwas  hashing  hellerstein  heritability  hinton  history  holdout  hypothesis-testing  ian-goodfellow  ideas  immune-system  immunology  importance-sampling  incentive-compatible-mechanism  incentives  inference  influence-function  information  instrumental-variable  interactivity  interpretation  iq  jan-draisma  jeff-heer  jmlr  joseph-halpern  kari-stefansson  kernel-methods  lasso  lda  learning-theory  learning-to-search  learning  linear-programming  linearization  local-differential-privacy  machine-learning  machinelearning  mapreduce  mark-gerstein  mark-girolami  markets  mathematics  matt-might  mcmc  mechanism-design  memory-allocation  memory  message-passing  metagenomics  metatheory  michael-jordan  michael-mitzenmacher  missing-data  model-checking  motion  msft-research  multi-party-computation  music  negation  network  networks  neural-networks  neuroscience  nicholas-kaldor  nips  online-algorithms  optimization  overfitting  p-hacking  parallel-computing  partitioning  paul-romer  perceptual-speed  persi-diaconis  peter-alvaro  peter-park  phenome  pitch-estimation  political-science  population-genetics  prediction  preprint  pricing  printing-press  privacy  probabilistic-methods  probabilistic-programming  probability  programming-languages  programming  programminglanguage  rafael-frongillo  randomization  randomized-experiments  reference-free  reference-graphs  region-join  regression  regret  religion  sampling  sanger-sequencing  security  sequence-analysis  sequencing  siggraph  sigma  signal-processing  similarity  sketching  social-media  social-networks  sociology  somatic-variation  spark  sparse-methods  spectral-clustering  sql  statistical-learning  statistical-significance  statistics  stick-breaking  streaming  string-processing  strings  suffix-arrays  survey  swan  tensor-decomposition  testing  topic-modeling  training-data  training  type-systems  ui  united-states  usage  vega  versioning  victor-veitch  vikash-mansinghka  visualization  wordvec  work  yee-whye-teh  yoshua-bengio  zig-zag 

Copy this bookmark: