《狼和羊的故事.ppt》由会员分享，可在线阅读，更多相关《狼和羊的故事.ppt（43页珍藏版）》请在三一办公上搜索。
1、第5章 学习LEARNING,Introduction to PsychologyDepart.Psycho.Peking University,Garcia et al(1985)狼和羊的故事,含氯化锂的羊肉汉堡包恶心羊进入狼的地盘约一小时之内，狼尝试多次，但无攻击,什么改变了狼的天性？,本能行为通过遗传获得，不学就会的行为。母鸡孵蛋、蜜蜂酿蜜、婴儿吸奶药物操作（手脚）学习！-习得行为人-习得行为和本能行为的混合体,本能行为,习得行为,Outlines,What is learning?Classical ConditioningOperant ConditioningCognitive-S
2、ocial Theory,1 What is learning?,1.1 Definition,Learning refers to any enduring change in the way an organism responds based on its experience.因为经验而使有机体的反应方式产生持久的改变疲劳、药物、疾病和损伤不属于学习范畴,Key Elements of Learning,A change in behavior or behavioral potentialA relatively permanent changeBased on experience
3、,1.2 Why is learning important?,The environment does not stand still分辨食物与生存环境天性有限（小孩喜欢甜味）The environment varies(Age,place,culture)袒胸露臂与迷你裙老实厚道与无能Human being-Social animal认知、情感、态度、人格、价值信仰等,1.3 Sharing assumptions,Experience shapes behavior经验塑造行为Learning is adaptive学习是适应Experimental research on the la
4、ws of learning学习可以用实验法进行研究,2 Classical Conditioning,巴甫洛夫（Ivan Pavlov）（1849-1936）研究狗的消化系统（Nobel winner）唾液分泌（Salivation）反射（reflex）华生（Watson）动物和人类适用同样的学习法则,2.1 Pavlov Model,Conditioning&Unconditioned reflex学习是一种联系的建立（获得）The organism learns to associate CS with UCSUCS,UCRCR,CRNS，中性刺激，与UCS多次重复出现.,Classic
5、al conditioning,食物,唾液分泌,铃声,定向探寻,UCS,UCR(CR),NS(CS),2.2 Conditioned Emotional Responses,触景生情一朝被蛇咬，十年怕井绳秋天的悲沧和凄凉歌曲月亮代表我的心爱在深秋把悲伤留给自己心太软,Watson&Rayner(1920)Little Albert,狗、兔子、白鼠、面具（圣诞老人）NS身后敲击金属棒 UCS跳起来、趴下、呜咽 UCR两月之后，小孩怕狗等（泛化）结论：情绪也是习得的,2.3 Conditioned Social Behavior,解释普遍存在的现象望梅止渴大鼠（胰岛素、葡萄糖）休克癌症化疗病人（食
6、物、护士、医院）免疫能力（糖精水）家庭、单位广告环境,2.4 Stimulus Generalization and Discrimination,新刺激越接近原刺激，越可能唤起反应 这种对类似刺激的反应就是泛化。分化(discrimination)，是对刺激差异的反应。,2.5 Extinction,CS不再与UCS一起出现，条件反应会减 弱，这就是消退。即刻恢复(spontaneous recovery)泛化、分化和消退的进化适应意义,2.6 Factors that Affect Classical Conditioning,Time schedule for CS&UCSForward
7、 conditioningSimultaneous conditioningBackward conditioningThe Individual Learning HistoryHigher-order conditioningBlocking effect(Kamin,1969)Preparedness to Learn（进化的角度）电击=避光、声，不避味道X光=不避光、声，避味道(Garcia&Koelling,1966),Contingency,3 Operant Conditioning,Edward L.Thorndike:Law of EffectS-R connection尝试
8、错误机制(blind trial-error)行为倾向取决于该行为产生的环境效果Law of effect=Instrumental conditioningB.F.Skinner:Operant Conditioning学习是一种反应概率上的变化，而强化是增强反应概率的手段。,3.1 Operants and Consequence,Operants:Emitted/ElicitedInstrumental conditioning(Thorndike)ConsequenceReinforcementPunishment,3.2 Reinforcement,Making the respon
9、se more likely to recurTwo kinds of ReinforcementPositive Reinforcement提供奖赏Negative Reinforcement终止一个负面刺激（负面强化物）,3.3 Punishment,Decreases the probability that a response will recurPositive&Negative PunishmentCommon Problem in Using Punishment,惩罚效果差的原因,惩罚与操作难以对号对惩罚人而非受罚行为害怕存在其他奖励惩罚受（领导）长辈的情绪所左右（一致性差）
10、攻击性惩罚导致更多的攻击行为 建议：惩罚说理（或强化）,3.4 Schedules of Reinforcement,Continuous reinforcement schedulePartial or Intermittent schedulesRatio schedulesFixed RatioVariable RatioInterval schedulesFixed IntervalVariable Interval,3.5 Factors Affecting Operant Conditioning,操作与奖惩之间的联系的偶然性Skinner(1948)鸽子的迷信行为操作与后果之间的
11、间隔特异刺激的呈现:stimulus control学习某种行为的准备性学习者的特点:antisocial personality,3.6 Why is reinforcer reinforcing?,Reinforcers as Drive ReducersC.Hull(1943):Drive-reduction theoryPrimary and Secondary ReinforcersConditioned reinforcersD.Premack(1965):Premack Principlepreferred activity as ReinforcementThe Role of
12、 FeelingsEmotion/sensory experiencesThe behavioral approach system(pleasure)The behavioral inhibition system(anxiety)注意：强化物的针对性和有效性。,3.7 The Reciprocity of Social Interactions,因果链相互操作,3.8 The Relation between Classical and Operant Conditioning,应答行为工具性行为诱发反应自发反应刺激在先反应在先特定的刺激无特定刺激自主NS躯体NS,3.9 The appl
13、ication of Operant Conditioning,N.Miller(1978):Biofeedback trainingBehavior shaping and ChainingProgrammed Instruction,The Procedure for Behavior shaping and Chaining,Reinforcing closer and closer approximations to the desired response.A sequence of already-established behaviors is reinforced step b
14、y step.行为塑造的要领确定最终目标对目标加以分解逐步强化分解后的目标动物驯养的常用技术,Programmed Instruction,机器教学（计算机辅助教学CAI）总目标分解为多个学习单元进一步分解为多个步骤及时反馈和强化学习效果可以自定学习路径和节奏,其他应用形式,奖励（劳模、三好生）代币制（红花、卡片）惩罚（交通处罚累积分制）,4 Cognitive-Social Theory,复杂的学习怎么解释？强化如何让人学习复杂的行为？强化是学习的必要元素吗？主体的作用,4.1 The Role of Cognition,Insight,Gestalt：完形学习理论柯勒：黑猩猩够香蕉顿悟而非
15、盲目尝试错误的紊乱动作内心构造完形整体大于各个部分之和,Cognitive Map,Edward C.Tolman(1948),刺激-反应,刺激-主体（意识）-反应,Latent Learning,Tolman&Honzik(1930)结论：学习不一定需要强化,Expectation and Locus of Control,Julian Rotter(1954，1990)内控（Internal locus of control）外控（External locus of control）,Learned helplessness and Explanatory style,Martin Sel
16、igman(1975)Learned helplessness 习得无助预期自己不能逃避不利的事情，放弃努力Explanatory style悲观主义（Pessimistic explanatory style）父母越悲观，子女成绩和幸福感越差(Belt&Peterson,1991),4.2 The Role of Social Learning,A.Bandura(1963)班图拉Social learningObservational learningModelingVicarious conditioning,结语,先天与后天情境与认知,Assignment No.3,生活中有人吸烟，有人不吸烟。请你根据自己的体会谈吸烟的人是如何染上吸烟习惯的？或者你为什么没有学会吸烟？吸烟的人为什么很难戒烟？,
