Go-directed and habitual learning

行为心理学

行为心理学由instrumental conditioning(operant conditioning)和respondent behaviors(classical conditioning; Pavlovian)共同构成。两者区别在于stimuli和response是voluntary 还是involuntary。
Classical conditioning 典型的代表是巴普洛夫的狗,中性条件(铃声)配合无条件刺激(食物)成为有条件刺激(狗听到铃声会分泌唾液).


Go-directed and habitual learning

instrumantal conditioning 是指通过奖励和惩罚来加强或者削弱未来这个动作会出现的概率。
Go-directed and habitual learning
instrumantal conditioning is governed by two forms of action control: go-directed and habit learning procss.

  • Go-directed learning contains medial prefrontal cortex, dorsomedial striatum(DMS) and thalamus + feedback to cortex.
  • habits includes sensory-motor cortex, dorsolateral striatum, thalamus and feedback to cortex.

They are in a competitive relationship, 一个环路损坏会导致另一回路的产生。可以用fMRI(functional maganetic resonance Imagining)来查看

通常有两种方式在行为上检验它们。

  • outcome devaluation
    this aims to evaluate the sensitivity to changes of outcome values(1. By making the outcome less attractive; 2. by pairing the outcome with awful things)

  • contigency degradation
    this aims to evaluate the sensitivity to changes in causal relationship between action and consequence (1. by replacing a delayed outcome 2. by disrupting the original causal relationship)

    Existing computational models

Go-directed and habitual learning

上一篇:Python接口测试中通过登录接口获取实时token


下一篇:【转】关于ARM Cortex内核的种类以及性能解读