A survey of machine learning机器学习分享综述.pdf

《A survey of machine learning机器学习分享综述.pdf》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、AsurveyofmachinelearningFirsteditionbyCarlBurchforthePennsylvaniaGovernor'sSchoolfortheSciencesc2001,CarlBurch.Allrightsreserved.iiContents0Introduction10.1Datamining...........................................10.2Neuralnetworks......................................

2、...20.3Reinforcementlearning.....................................20.4Artificiallife...........................................21Datamining31.1Predictingfromexamples....................................31.2Techniques...........................................41.2.1Line

3、arregression....................................51.2.2Nearestneighbor....................................81.2.3ID3...........................................91.3Generalissues..........................................131.4Conclusion...................................

4、........142Neuralnetworks152.1Perceptron............................................152.1.1Theperceptronalgorithm................................152.1.2Analysis........................................182.2Artificialneuralnetworks....................................19

5、2.2.1ANNarchitecture....................................192.2.2Thesigmoidunit....................................202.2.3Predictionandtraining.................................212.2.4Example........................................232.2.5Analysis.......................

6、.................253Reinforcementlearning273.1Modelingtheenvironment...................................273.2
learning............................................283.3Nondeterministicworlds....................................323.4Fastadaptation.......................

7、..................323.5TDlearning...........................................343.6Applications...........................................343.6.1TD-Gammon......................................343.6.2Elevatorscheduling...................................354References37i

8、vCONTENTSChapter0IntroductionMachinelearningattemptstotellhowtoautomaticallyfindagoodpredictorbasedonpastexperiences.Althoughyoumightarguet