ai   79715

« earlier    

Welcome to the Institute for Artificial Intelligence! [Artificial Intelligence]
The Institute for Artificial Intelligence (IAI) directed by Prof. Michael Beetz is part of the Faculty of Computer Science and member of the Center for Computing and Communication Technologies (TZI) at the University of Bremen.
ai  research  Deutschland  robotics 
yesterday by mandarine
Teachable Machine
Explore machine learning, live in your browser.
ai  games 
yesterday by hanyu
Competitive Self-Play
Self-play allows simulated AIs to discover physical skills like tackling, ducking, faking, kicking, catching, and diving for the ball, without explicitly designing an environment with these skills in mind.
yesterday by hanyu
What Happens When Robots Act Just Like Humans
Ishi­guro believes that since we’re hardwired to interact with and place our faith in humans, the more humanlike we can make a robot appear, the more open we’ll be to sharing our lives with it. Toward this end, his teams are pioneering a young field of research called human-robot interaction.

HRI is a hybrid discipline: part engineering, part AI, part social psychology and cognitive science. The aim is to analyze and cultivate our evolving relationship with robots. HRI seeks to understand why and when we’re willing to interact with, and maybe even feel affection for, a machine. And with each android he produces, Ishiguro believes he is moving closer to building that trust.

As complex as we assume ourselves to be, our bonds with one another are often built on very ­little. Given all the time we now spend living through technology, not many of us would notice, at least at first, if the friend we were messaging were replaced by a bot. And humans do not require much to stir up feelings of empathy with another person or creature—even an object. In 2011 a University of Calgary test found that subjects were quick to assign emotions and intentions to a piece of balsa wood operated with a joystick. In other words, we are so hardwired for empathy that our brains are willing to make the leap to humanizing a piece of wood. It’s a level of animal instinct that’s slapstick-hilarious and a degree of vulnerability that’s terrifying.
AI  Longform  robotics 
yesterday by jbertsche
Machine Learning FAQ
Random Forests vs. SVMs

I would say that random forests are probably THE “worry-free” approach - if such a thing exists in ML: There are no real hyperparameters to tune (maybe except for the number of trees; typically, the more trees we have the better). On the contrary, there are a lot of knobs to be turned in SVMs: Choosing the “right” kernel, regularization penalties, the slack variable, …

Both random forests and SVMs are non-parametric models (i.e., the complexity grows as the number of training samples increases). Training a non-parametric model can thus be more expensive, computationally, compared to a generalized linear model, for example. The more trees we have, the more expensive it is to build a random forest. Also, we can end up with a lot of support vectors in SVMs; in the worst-case scenario, we have as many support vectors as we have samples in the training set. Although, there are multi-class SVMs, the typical implementation for mult-class classification is One-vs.-All; thus, we have to train an SVM for each class – in contrast, decision trees or random forests, which can handle multiple classes out of the box.

To summarize, random forests are much simpler to train for a practitioner; it’s easier to find a good, robust model. The complexity of a random forest grows with the number of trees in the forest, and the number of training samples we have. In SVMs, we typically need to do a fair amount of parameter tuning, and in addition to that, the computational cost grows linearly with the number of classes as well.
ai  howto  algorithm 
yesterday by janpeuker
Research Blog: TensorFlow Lattice: Flexibility Empowered by Prior Knowledge
We take advantage of the look-up table’s structure, which can be keyed by multiple inputs to approximate an arbitrarily flexible relationship, to satisfy monotonic relationships that you specify in order to generalize better. That is, the look-up table values are trained to minimize the loss on the training examples, but in addition, adjacent values in the look-up table are constrained to increase along given directions of the input space, which makes the model outputs increase in those directions
ai  google  library  analytics 
yesterday by janpeuker

« earlier    

related tags

#2017  #ai  #competition  #impact  #outlook  #report  #significance  #uk  10  addiction  aiux  algorithm  algorithmic_transparency  amazon  analytics  and  andrewng  apac  api  artificialintelligence  auditing  automation  bezels  bias  bigdata  blockchain  books  boston  brain  chatbot  choices  cocacola  code  company  comparison  computer  content  conversational.interfaces  conversational  course  cybersecurity  data  data_science  datamining  datascience  decision  deep-learning  deep  deeplearning  defstar5  design  deutschland  digital  digitalmarketing  dl  ecommerce  edtech  ethics  experiment  eye  facebook  fakenews  features  fintech  framework  future  game  games  genetic-algorithms  google  growthhacking  gwern  hands-on:  hcds  hcm  history  home  how-to  howto  huawei  identity  ii  iiot  image  inboundmarketing  industry40  innovation  insurtech  iot  jm  kindle  landscape  learning  library  linkedin  longform  m2m  machine-learning  machine  machine_learning  machinelearning  machinlearning  makeyourownlane  manufacturing  market_research  marketi  marketing  mate  measurement  meta_learning  microsoft  mind  ml  mpgvip  neural-networks  organization  perceptron  pharma  platforms  prediction  programming  psychology  python  ranking  read  recognition  report  research  research_ethics  reward-hacking  robotics  science  sckit  self-learning  shipping  slow  smarts  social  socialmedia  statistics  storage  sw  tech  technology  technology_conference  telenorntnuai  testing  text  thinking  tiny  training  transparency  trees  trends2018  uk_govt  uka  usa  ux  vc  weird  white_paper  workplace   

Copy this bookmark: