auto-learning   14

Cultural variation in cultural evolution | Proceedings of the Royal Society of London B: Biological Sciences
Cultural evolutionary models have identified a range of conditions under which social learning (copying others) is predicted to be adaptive relative to asocial learning (learning on one's own), particularly in humans where socially learned information can accumulate over successive generations. However, cultural evolution and behavioural economics experiments have consistently shown apparently maladaptive under-utilization of social information in Western populations. Here we provide experimental evidence of cultural variation in people's use of social learning, potentially explaining this mismatch. People in mainland China showed significantly more social learning than British people in an artefact-design task designed to assess the adaptiveness of social information use. People in Hong Kong, and Chinese immigrants in the UK, resembled British people in their social information use, suggesting a recent shift in these groups from social to asocial learning due to exposure to Western culture. Finally, Chinese mainland participants responded less than other participants to increased environmental change within the task. Our results suggest that learning strategies in humans are culturally variable and not genetically fixed, necessitating the study of the β€˜social learning of social learning strategies' whereby the dynamics of cultural evolution are responsive to social processes, such as migration, education and globalization.


Western education emphasizes individual discovery and creativity, whereas East Asian education emphasizes rote learning from authority [25]. The adoption of consumer products shows less social influence in Western than East Asian countries [26]. Westerners are described as more individualistic/independent, while East Asians are described as more collectivistic/interdependent [27], dimensions which intuitively map on to asocial and social learning, respectively.

Societal background influences social learning in cooperative decision making:
We demonstrate that Chinese participants base their cooperation decisions on information about their peers much more frequently than their British counterparts. Moreover, our results reveal remarkable societal differences in the type of peer information people consider. In contrast to the consensus view, Chinese participants tend to be substantially less majority-oriented than the British. While Chinese participants are inclined to adopt peer behavior that leads to higher payoffs, British participants tend to cooperate only if sufficiently many peers do so too. These results indicate that the basic processes underlying social transmission are not universal; rather, they vary with cultural conditions. As success-based learning is associated with selfish behavior and majority-based learning can help foster cooperation, our study suggests that in different societies social learning can play diverging roles in the emergence and maintenance of cooperation.
study  org:nat  anthropology  cultural-dynamics  sapiens  pop-diff  comparison  sociality  learning  duplication  individualism-collectivism  n-factor  europe  the-great-west-whale  china  asia  sinosphere  britain  anglosphere  strategy  environmental-effects  biodet  within-without  auto-learning  tribalism  things  broad-econ  psychology  cog-psych  social-psych  🎩  🌞  microfoundations  egalitarianism-hierarchy  innovation  creative  explanans  education  culture  curiosity  multi  occident  cooperate-defect  coordination  organizing  self-interest  altruism  patho-altruism  orient  ecology  axelrod 
may 2018 by nhaliday
'Poisoning Attacks against Support Vector Machines', Battista Biggio, Blaine Nelson, Pavel Laskov
The perils of auto-training SVMs on unvetted input.
We investigate a family of poisoning attacks against Support Vector Machines (SVM). Such attacks inject specially crafted training data that increases the SVM's test error. Central to the motivation for these attacks is the fact that most learning algorithms assume that their training data comes from a natural or well-behaved distribution. However, this assumption does not generally hold in security-sensitive settings. As we demonstrate, an intelligent adversary can, to some extent, predict the change of the SVM's decision function due to malicious input and use this ability to construct malicious data. The proposed attack uses a gradient ascent strategy in which the gradient is computed based on properties of the SVM's optimal solution. This method can be kernelized and enables the attack to be constructed in the input space even for non-linear kernels. We experimentally demonstrate that our gradient ascent procedure reliably identifies good local maxima of the non-convex validation error surface, which significantly increases the classifier's test error.

Via Alexandre Dulaunoy
papers  svm  machine-learning  poisoning  auto-learning  security  via:adulau 
july 2012 by jm

related tags

academia  acm  acmtariat  ai  altruism  analogy  anglosphere  announcement  anthropology  applications  asia  atoms  automation  axelrod  bayesian  biodet  britain  broad-econ  china  cog-psych  commentary  comparison  concept  cooperate-defect  coordination  creative  cultural-dynamics  culture  curiosity  data-science  data  deep-learning  deepgoog  defense  detail-architecture  duplication  ecology  education  egalitarianism-hierarchy  engineering  environmental-effects  europe  explanans  fixed-point  flexibility  games  generalization  google  gradient-descent  hn  homepage  individualism-collectivism  innovation  interface  learning  libraries  liner-notes  machine-learning  microfoundations  military  multi  n-factor  neuro  news  nibble  occident  org:bleg  org:mat  org:nat  org:sci  organizing  orient  papers  patho-altruism  poisoning  pop-diff  preprint  programming  project  psychology  python  reinforcement  research-program  research  sapiens  security  self-interest  sequential  sinosphere  skunkworks  social-psych  sociality  speedometer  state-of-art  stats  strategy  study  summary  svm  tech  techtariat  the-great-west-whale  things  time-series  tribalism  virginia-dc  wire-guided  within-without  yak-shaving  🌞  🎩  πŸ–₯ 

Copy this bookmark: