machine-learning   23212

« earlier    

Is attacking machine learning easier than defending it?
So far, it is much easier to design tricks like this that fool a model than it is to design models that can’t be fooled.

Adversarial examples are hard to defend against because they require machine learning models to produce good outputs for every possible input. Most of the time, machine learning models work very well but only work on a very small amount of all the many possible inputs they might encounter.
machine-learning 
yesterday by danmichaelo
Artificial Intelligence — The Revolution Hasn’t Happened Yet
There is a different narrative that one can tell about the current era. Consider the following story, which involves humans, computers, data and life-or-death decisions, but where the focus is something other than intelligence-in-silicon fantasies. When my spouse was pregnant 14 years ago, we had an ultrasound. There was a geneticist in the room, and she pointed out some white spots around the heart of the fetus. “Those are markers for Down syndrome,” she noted, “and your risk has now gone up to 1 in 20.” She further let us know that we could learn whether the fetus in fact had the genetic modification underlying Down syndrome via an amniocentesis. But amniocentesis was risky — the risk of killing the fetus during the procedure was roughly 1 in 300. Being a statistician, I determined to find out where these numbers were coming from. To cut a long story short, I discovered that a statistical analysis had been done a decade previously in the UK, where these white spots, which reflect calcium buildup, were indeed established as a predictor of Down syndrome. But I also noticed that the imaging machine used in our test had a few hundred more pixels per square inch than the machine used in the UK study. I went back to tell the geneticist that I believed that the white spots were likely false positives — that they were literally “white noise.” She said “Ah, that explains why we started seeing an uptick in Down syndrome diagnoses a few years ago; it’s when the new machine arrived.”
machine-learning  stats  argument  politics  AI  life  reference 
yesterday by kmt
AI software that helps doctors diagnose like specialists is approved by FDA - The Verge
For the first time, the US Food and Drug Administration has approved an artificial intelligence diagnostic device that doesn’t need a specialized doctor to interpret the results. The software program, called IDx-DR, can detect a form of eye disease by looking at photos of the retina.
machine-learning  med 
yesterday by tamberg

« earlier    

related tags

acmtariat  add-to-list  ai-control  ai-policy  ai  algo  algorithmism  algorithms  alignment  analogy  api  apple  archive  argument  art  articles  artificial-intelligence  arxiv  atoms  audio-generation  audio  auto-learning  bias  blog  browser  business  calendar  chaos  cheatsheet  citizen-science  classification  clever-rats  computer-architecture  computer-vision  concept  content-samurai  course  courses  crypto  data-analysis  data-science  data  datascience  deductive  deep-learning-compiler  deep-learning  deepgoog  deeplearning  deepmind  design  dev  development  django  dnn  docker  email  engineering  ethics  executive-summary  explanation  exposition  flask  flexibility  future  games  gender  github  google  gradient-descent  graphics  hack  heard-the-talk  history  howto  human-ml  humanism  humanities  image-generation  image-recognition  image  imaging  inductive  intel  iot  iteration-recursion  javascript  learning  lecture  lectures  libraries  library  life  liner-notes  lstm  materials  math  mathematics  med  medium  michael-i-jordan  ml  multimedia  networks  neural-net  neural-network  neuro  nibble  nlp  open-ai  optimization  org:bleg  p5js  papers  parallel  physics  pocket  politics  prediction  probability  programming  project  python  r-project  ratty  reference  reinforcement-learning  reinforcement  research-program  research  retail  robotics  safari  scala  science  scienceontheweb  search  security  siri  spark  speech-generation  sql  stability  statistics  stats  study-group  summary  systems  technology  techtariat  tensorflow  tensorflowjs  tesla  text-extraction  text  threads  time-series  time  to-read  tradeoffs  transportation  turi-create  tutorial  tutorials  values  vector-embedding  video  videogame  vision  voice  volo-avolo  vowpal-wabbit  voxable  vw  waymo  webgl  wire-guided  wow 

Copy this bookmark:



description:


tags: