資源描述:
《概率論和數(shù)理統(tǒng)計(jì)隨機(jī)變量的數(shù)字特征.ppt》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在教育資源-天天文庫。
1、1§1.3隨機(jī)變量的數(shù)字特征一、數(shù)學(xué)期望與方差二、協(xié)方差與協(xié)方差Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.2若當(dāng)級數(shù)絕對收斂時(shí),稱為隨機(jī)變量X的數(shù)學(xué)期望,記為E(X),即Xx1x2x3………xn…Pkp1p2p3………pn…1、數(shù)學(xué)期望的定義定義2設(shè)連續(xù)型隨機(jī)變量X的概率密度為f(x),則當(dāng)廣義積分絕對收斂時(shí),稱此積分的值為隨機(jī)變量X的數(shù)學(xué)期望,記為E(X),即E(X)=E(X)=一、數(shù)學(xué)期望與方差1、定義1設(shè)
2、離散型隨機(jī)變量X的分布律為:Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.32、數(shù)學(xué)期望的性質(zhì):(4)若X,Y為兩個(gè)相互獨(dú)立的隨機(jī)變量,則有E(XY)=E(X)E(Y)(1)設(shè)C是常數(shù),則E(C)=C這里C視為退化的隨機(jī)變量(2)設(shè)X為一隨機(jī)變量,C為常數(shù),則有E(CX)=CE(X)(3)設(shè)X,Y為兩個(gè)隨機(jī)變量,則有E(X+Y)=E(X)+E(Y)注:(1)相互獨(dú)立時(shí)(2)Evaluationonly.Create
3、dwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.4例2、已知X~E(X),求Y=2X-1的數(shù)學(xué)期望解依題意知,X的概率密度為于是進(jìn)而Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.3、隨機(jī)變量函數(shù)的數(shù)學(xué)期望⑴離散型:X的分布率為:P{X=xk}=Pk,k=1,2…且級數(shù)5⑵連續(xù)型:X的概率密度為f(x),若
4、積分(1)已知隨機(jī)變量X的分布,求其函數(shù)Y=g(X)的期望:絕對收斂絕對收斂Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.6(2)連續(xù)型R.V(X,Y)的概率密度為:f(x,y)則有(1)離散型(X,Y)的分布律為:(2)、已知隨機(jī)變量(X,Y)的分布,求函數(shù)Z=g(X,Y)的數(shù)學(xué)期望求的期望例3:已知隨機(jī)變量X的概率密度為Evaluationonly.CreatedwithAspose.Slidesfor.NET
5、3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.7例1.26設(shè)隨機(jī)變量解依題知,X的概率密度為故Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.4、方差的概念8另外,記,稱為標(biāo)準(zhǔn)差或均方差D(x)=Var(X)=存在,則稱之為X的方差.記為D(X)或Var(X)定義若X是一隨機(jī)變量,若5、方差的計(jì)算方法:當(dāng)X為離散型隨機(jī)變當(dāng)X是連續(xù)型隨機(jī)變量常用公式:Ev
6、aluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.9例5:已知X~U(a,b),求E(X)和D(X).解由題知,X的概率密度為于是有而Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.6、方差的性質(zhì):10(1)D(C)=0;(2)D(CX)=C2D(X);(3)當(dāng)X、Y獨(dú)立
7、,D(X+Y)=D(X)+D(Y);(4)D(X)=0等價(jià)于P﹛X=C﹜=1.(C為常數(shù))Evaluationonly.CreatedwithAspose.Slidesfor.NET3.5ClientProfile5.2.0.0.Copyright2004-2011AsposePtyLtd.7、常見分布的期望方差:11(5)均勻分布:(1)二點(diǎn)分布:(2)二項(xiàng)分布:(3)泊松分布:(4)正態(tài)分布:E(X)=npD(X)=np(1-p)(6)指數(shù)分布E(X)=pD(X)=pqEvaluationonly.CreatedwithAspose.Slides