Question

Time series analysis. What are the two fundamental assumptions in time series analysis? Explain in words (without writing down equations). What are the different steps in the generation of a time series using an MA(2) model. Explain just in words, the only equation that you should write down is the general equation for an MA(2) model. There is no need to write down other equations.

Answer 1

Jiod X1 time seaies are The two frodameotal assumptions iowestibility Stationasity 0/2 కలు Shationas itg aquiaument main Csitea ia One af the the thu fosucasts ou ton a time Series data eithes has mo 3F a 8Cwes not 8tationaag means, that the Sesies vaslance auto co vasiane moottoni audependent Constamt mean On time no fs said to be Stationasy 4 thue is e, A tima Sies no sgstematic stema Chage MeanCi no tund) the is omd Staictly Pesiodce have vasiatcons change in vasiance PRopestces of One Section of been wasds the Removed the those af much data o the seetcon. n osdes to ensus th uniqcan of model inveatblit Condition must be needed f Moving avesase model com autougesive medul iplies h invextiblity be Whitten as an we Kncw -hat AR model is alway iovestible, and UA model s alway Stalionay Jova tibi slet autocomelation ncton of moving auesage Phocess non uniquenuss of

MAC2) PaocesS said to be mo vi aveiage Paocers af A time sesie t is lincas Sum t Osdes q, abbsival ed the as MAC2) IS of past 2 values af Random So that Shocks e a puely handom PAOcess X ,B-0,8) with and Constomt Raso mem Vasiance B backwasd hift opesato ie, BXtXt- ;e cB)t-18- 0,B2 Di4feunt Steps O Plot the clata txend To cheek o hethes -lhere exist of differuncing Deju mean Stationasy, thun difference the data t4 dlata s not have mean Stetionay diffeunced data unit -the iduntificafcon of DR MO ORdes Plotng AcF amd PACF Test the detesministfc Frend taam Selee tion Model usin minimum AIC and BIc Diasnostrc Cheking Model as sump tions au satis-fied os nd- Chick whithes using histogaam, PPAlot on Q- plot, Residual plot Forucast