jm + models   9

Applied machine learning at Facebook: a datacenter infrastructure perspective
Lots of cool details into how they've productized and scaled up their prod ML infrastructure.
As we looked at last month with Continuum, the latency of incorporating the latest data into the models is also really important. There’s a nice section of this paper where the authors study the impact of losing the ability to train models for a period of time and have to serve requests from stale models. The Community Integrity team for example rely on frequently trained models to keep up with the ever changing ways adversaries try to bypass Facebook’s protections and show objectionable content to users. Here training iterations take on the order of days. Even more dependent on the incorporation of recent data into models is the news feed ranking. “Stale News Feed models have a measurable impact on quality.” And if we look at the very core of the business, the Ads Ranking models, “we learned that the impact of leveraging a stale ML model is measured in hours. In other words, using a one-day-old model is measurably worse than using a one-hour old model.” One of the conclusions in this section of the paper is that disaster recovery / high availability for training workloads is key importance.
machine-learning  facebook  ml  training  ops  models  infrastructure  prod  production 
23 hours ago by jm
3D models by DH_Age Sheela-na-Gig3D Project (@DH_Age) - Sketchfab
These are fantastic -- 3D scans of Sheela-na-Gig carvings around Ireland from 3D Sheela, an Irish based research initiative 'focusing on the digital documentation and analysis of Ireland's Sheela-na-Gig catalogue' (NSFW)
3d  sheela-na-gigs  history  carving  nsfw  models  photogrammetry 
13 days ago by jm
Brutal London
'A book about London's gorgeous, brutalist architecture includes dainty DIY papercraft models to make yourself' -- awesome
brutalist  architecture  london  papercraft  models  barbican 
november 2017 by jm
Fooling Neural Networks in the Physical World with 3D Adversarial Objects · labsix
This is amazingly weird stuff. Fooling NNs with adversarial objects:
Here is a 3D-printed turtle that is classified at every viewpoint as a “rifle” by Google’s InceptionV3 image classifier, whereas the unperturbed turtle is consistently classified as “turtle”.

We do this using a new algorithm for reliably producing adversarial examples that cause targeted misclassification under transformations like blur, rotation, zoom, or translation, and we use it to generate both 2D printouts and 3D models that fool a standard neural network at any angle. Our process works for arbitrary 3D models - not just turtles! We also made a baseball that classifies as an espresso at every angle! The examples still fool the neural network when we put them in front of semantically relevant backgrounds; for example, you’d never see a rifle underwater, or an espresso in a baseball mitt.
ai  deep-learning  3d-printing  objects  security  hacking  rifles  models  turtles  adversarial-classification  classification  google  inceptionv3  images  image-classification 
november 2017 by jm
Research Blog: Federated Learning: Collaborative Machine Learning without Centralized Training Data
Great stuff from Google - this is really nifty stuff for large-scale privacy-preserving machine learning usage:

It works like this: your device downloads the current model, improves it by learning from data on your phone, and then summarizes the changes as a small focused update. Only this update to the model is sent to the cloud, using encrypted communication, where it is immediately averaged with other user updates to improve the shared model. All the training data remains on your device, and no individual updates are stored in the cloud.

Federated Learning allows for smarter models, lower latency, and less power consumption, all while ensuring privacy. And this approach has another immediate benefit: in addition to providing an update to the shared model, the improved model on your phone can also be used immediately, powering experiences personalized by the way you use your phone.

Papers:
https://arxiv.org/pdf/1602.05629.pdf , https://arxiv.org/pdf/1610.05492.pdf
google  ml  machine-learning  training  federated-learning  gboard  models  privacy  data-privacy  data-protection 
april 2017 by jm
Combining static model checking with dynamic enforcement using the Statecall Policy Language
This looks quite nice -- a model-checker "for regular programmers". Example model for ping(1):

<pre>01 automaton ping (int max_count, int count, bool can_timeout) {
02 Initialize;
03 during {
04 count = 0;
05 do {
06 Transmit_Ping;
07 either {
08 Receive_Ping;
09 } or (can_timeout) {
10 Timeout_Ping;
11 };
12 count = count + 1;
13 } until (count &gt;= max_count);
14 } handle {
15 SIGINFO;
16 Print_Summary;
17 };</pre>
ping  model-checking  models  formal-methods  verification  static  dynamic  coding  debugging  testing  distcomp  papers 
march 2015 by jm
CausalImpact: A new open-source package for estimating causal effects in time series
How can we measure the number of additional clicks or sales that an AdWords campaign generated? How can we estimate the impact of a new feature on app downloads? How do we compare the effectiveness of publicity across countries?

In principle, all of these questions can be answered through causal inference.

In practice, estimating a causal effect accurately is hard, especially when a randomised experiment is not available. One approach we've been developing at Google is based on Bayesian structural time-series models. We use these models to construct a synthetic control — what would have happened to our outcome metric in the absence of the intervention. This approach makes it possible to estimate the causal effect that can be attributed to the intervention, as well as its evolution over time.

We've been testing and applying structural time-series models for some time at Google. For example, we've used them to better understand the effectiveness of advertising campaigns and work out their return on investment. We've also applied the models to settings where a randomised experiment was available, to check how similar our effect estimates would have been without an experimental control.

Today, we're excited to announce the release of CausalImpact, an open-source R package that makes causal analyses simple and fast. With its release, all of our advertisers and users will be able to use the same powerful methods for estimating causal effects that we've been using ourselves.

Our main motivation behind creating the package has been to find a better way of measuring the impact of ad campaigns on outcomes. However, the CausalImpact package could be used for many other applications involving causal inference. Examples include problems found in economics, epidemiology, or the political and social sciences.
causal-inference  r  google  time-series  models  bayes  adwords  advertising  statistics  estimation  metrics 
september 2014 by jm
Sweden Solar System
the world's largest permanent scale model of the Solar System. The Sun is represented by the Ericsson Globe in Stockholm, the largest hemispherical building in the world. The inner planets can also be found in Stockholm but the outer planets are situated northward in other cities along the Baltic Sea. The system was started by Nils Brenning and Gösta Gahm and is on the scale of 1:20 million.


(via JK)
scale  models  solar-system  astronomy  sun  sweden  science  cool  via:jk 
august 2014 by jm

Copy this bookmark:



description:


tags: