斯坦福大学机器学习课程讲义—— 多变量的线性回归模型表达.pptx

斯坦福大学机器学习课程讲义—— 多变量的线性回归模型表达会计学第2页/共40页Multiple features (variables).Size (feet2)Number of bedroomsNumber of floorsAge of home (years)Price ($1000)2104514546014163240232153432303158522136178……………第3页/共40页Multiple features (variables).Size (feet2)Number of bedroomsNumber of floorsAge of home (years)Price ($1000)2104514546014163240232153432303158522136178……………Notation:= number of features= input (features) of training example.= value of feature in training example.第4页/共40页Hypothesis: Previously:第5页/共40页For convenience of notation, define .Multivariate linear regression.第6页/共40页第7页/共40页Linear Regression with multiple variablesGradient descent for multiple variablesMachine Learning第8页/共40页Hypothesis:Parameters:Cost function:Gradient descent:Repeat(simultaneously update for every )第9页/共40页New algorithm :Gradient DescentRepeatPreviously (n=1):Repeat(simultaneously update for )(simultaneously update )第10页/共40页第11页/共40页Linear Regression with multiple variablesGradient descent in practice I: Feature ScalingMachine Learning第12页/共40页Feature ScalingIdea: Make sure features are on a similar scale.size (feet2)E.g.= size (0-2000 feet2) = number of bedrooms (1-5)number of bedrooms第13页/共40页Feature ScalingGet every feature into approximately a range.第14页/共40页Mean normalizationReplace with to make features have approximately zero mean (Do not apply to ).E.g. 第15页/共40页第16页/共40页Linear Regression with multiple variablesGradient descent in practice II: Learning rateMachine Learning第17页/共40页Gradient descent“Debugging”: How to make sure gradient descent is working correctly.How to choose learning rate .第18页/共40页Making sure gradient descent is working correctly.Example automatic convergence test:Declare convergence if decreases by less than in one iteration.No. of iterations第19页/共40页Making sure gradient descent is working correctly.Gradient descent not worki