Building Statistical Forecast Models.ppt

《Building Statistical Forecast Models.ppt》由会员分享，可在线阅读，更多相关《Building Statistical Forecast Models.ppt（22页珍藏版）》请在麦多课文档分享上搜索。

1、Building Statistical Forecast Models,Wes Wilson MIT Lincoln Laboratory April, 2001,Experiential Forecasting,Idea: Base Forecast on observed outcomes in previous similar situations (training data) Possible ways to evaluate and condense the training data Categorization Seek comparable cases, usually e

2、xpert-based Statistical Correlation and significance analysis Fuzzy Logic Combines Expert and Statistical analysis Belief: Incremental changes in predictors relate to incremental changes in the predictand Issues Requirements on the Training Data Development Methodology Automation,Outline,Regression-

3、based Models Predictor Selection Data Quality and Clustering Measuring Success An Example,Statistical Forecast Models,Multi-Linear Regression F = w0 + S wi Piwi = Predictor Weightingw0 = Conditional Climatology Mean Predictor ValuesGAM: Generalized Additive Models F = w0 + S wi fi(Pi) fi = Structure

4、 Function, determined during regressionPGAM: Pre-scaled Generalized Additive Models F = w0 + S wi fi(Pi) fi = Structure Function, determined prior to regression The constant term w0 is conditional climatology less the weighted mean bias of the scaled predictors,Models Based on Regression,Training Da

5、ta for one predictor P vector of predictor values E vector of observed events Residual R2 = | FP E |2 Regression solutions are obtained by adjusting the parametric description of the forecast model (parameters w) until the objective function J(w) = R2 is minimized Multi-Linear Regression (MLR) J(w)

6、= | Aw E |2 MLR is solved by matrix algebra; the most stable solution is provided by the SVD decomposition of A,Regression and Correlation,Training Data for one predictor P vector of predictor values E vector of observed events Error Residual: R2 = | FP E |2 Correlation Coefficient r(P, E) = DP DE /

7、 sDPsDE Fundamental Relationship. Let F0 be a forecast equation with error residuals E0 (|E0|=R0). Let W0 + W1 P be a BLUE correction for E0, and let F = F0 + E0 . The error residual RF of F satisfiesRF2 = R02 1 - r(P, E0)2 ,Model Training Considerations,Assumption: The training data are representat

8、ive of what is expected during the implementation period Simple models are less likely to capture undesirable (non-stationary) short-term fluctuations in the training data The climatology of the training period should match that expected in the intended implementation period (decade scale) It is irr

9、ational to expect that short training periods can lead to models with long-term skill Plan for repeated model tuning Design self-tuning into the system It is desirable to have many more training cases than model parameters,The only way to prepare for the future is to prepare to be surprised; that do

10、esnt mean we have to be flabbergasted. Kenneth Boulding,GAM,An established statistical technique, which uses the training data to define nonlinear scaling of the predictors Standard implementation represents the structure functions as B-splines with many knots, which requires the use of a large set

11、of training data The forecast equations are determined by linear regression including the nonlinear scaling of the predictors F = w0 + Si wi fi(Pi) The objective is to minimize the error residual The structure functions are influence by all of the predictors, and may change if the predictor mix is a

12、ltered If a GAM model has p predictors and k knots per structure function, then the regression model has np+1 (linear) regression parameters,PGAM: Pre-scaled GAM,A new statistical technique, which permits the use of training sets that are decidedly smaller than those for GAM Once the structure funct

13、ions are selected, the forecast equations are determined by linear regression of the pre-scaled predictors F = w0 + S wi fi(Pi) Determination of the structure functions is based on enhancing the correlation of the (scaled) predictor with the error residual of conditional climatologyMaximize r( fi(Pi

14、), DE ) The structure function is determined for each predictor separately Composite predictors should be scaled as composites The structure functions often have interpretations in terms of scientific principles and forecasting techniques,Predictors,Every Method Involves a Choice of Predictors The G

15、reat Predictor Set: Everything relevant and available Possible Reduction based on Correlation Analysis Predictor Selection Strategies Sequential Addition Sequential Deletion Ensemble Decision ( SVD ) Changing the predictor list changes the model weights; for GAM, it also changes the structure functi

16、ons,Computing Solutions for the Basic Regression Problem,Setting: Predictor List Pi n and observed outcomes b over the m trials of the training set Basic Linear Regression ProblemA w = bwhere the columns of the m by n matrix A are the lists of observed predictor values over the trials Normal Equatio

17、ns: ATA w = ATb Linear Algebra: w = (ATA)-1 Atb Optimization: Find x to minimize R2 = | Aw b |2,SVD Singular Value Decomposition,A = U S VT where U and V are orthogonal matricesand S = S | 0 T where S is diagonal with positive diagonal entries UT A w = S VT w = UT b Set w = VTw, b = UTbn Restatement

18、 of the Basic Problem S VT w = b or S w = b (original problem space) (VT-transformed problem space) Since U is orthogonal, the error residual is not altered by this restatement of the problem,CAUTION: Analysis of Residuals can be misleading unless the dynamic ranges of the predictor values have been

19、 standardized,Structure of the Error Residual Vector,sis are usually decreasing sn 0, or reduce predictor list For i n, there is no solution. This is the portion of the problem that is not resolved by these predictors Magnitude of the unresolved portion of the problem: . R*2 = Sn+1m bi2,s1s2s3*sn,=,

20、w1 w2 w3* wn,b1 b2 b3* bnbn+1* bm,S w = b,Controlling Predictor Selection,SVD / PC analysis provides guidance Truncation in w space reduces the degrees of freedom Truncation does not provide nulling of predictors: . since 0 components of w . do not lead to 0 components of w = V w Seek a linear forec

21、ast model of the form F( a ) = aT w = S wi ai , a is a vector of predictor values Predictor Nulling: The ith predictor is eliminated from the problem if wi = 0 Benefits of predictor nulling Provides simple models Eliminate designated predictors (missing data problem) Quantifies the incremental benef

22、it provided by essential predictors (sensor benefit problem),Predictor Selection Process,Gross Predictor Selection (availability exhaustive searches are probably feasible, cross validation is wise.,Creating 15z Satellite Forecast Models (1),149 marine stratus days from 1996 to 2000 51 sectors and 3

23、potential predictors per sector (153) Compute the correlation for each predictor with the residual from conditional climatologyRetain only predictors, which have correlation greater than .25, reduces the predictor list to 45 predictors Separate analysis for two data sets, Raw and PGAM Truncate each

24、when SD reduction drops below 1.5 %,RAW:,PGAM:,Creating 15z Satellite Forecast Models (2),SVD Truncate 6 Pred.Nulling In the Truncation space: Null to 7 predictors with acceptable error growth Maximal Problems (R-8,P-7) Minimal Problems (R-5,P-4) Neither problem would accept augmentation according t

25、o the strict cross-validation test Different predictors were selected,Data Quality and Clustering,DQA is similar to NWP need to do the training set probably need to work to tighter standards Data Clustering During training - manual + For implementation - fully automated Conditional Climatology based

26、 on Clustering,Satellite Statistical Model (MIT/LL),1-km visible channel (brightness) Data pre-processing re-mapping to 2 km grid 3x3 median smoother normalized for sun angle calibrated for lens graying Grid points grouped into sectors topography physical forcing operational areas Sector statistics

27、Brightness Coverage Texture 4 year data archive, 153 predictors PGAM Regression Analysis,SECTORIZATION,Consensus Forecast,Satellite SFM,Regional SFM,Local SFM,COBEL,Forecast Weighting Function,Day Characterization- Wind direction- Inversion height- Forcing influences,Consensus Forecast,Measuring Suc

28、cess,Conclusions,PGAM, SVD/PC, and Predictor Nulling provides a systematic way to approach the development of Linear Forecast models via Regression This methodology provides a way to investigate the elimination of specific predictors, which could be useful in the development of contingency models We are investigating full automation,