Bootstrap

matlab使用随机傅里叶特征(Random Fourier Features, RFF)进行递归的样本外(Out-of-Sample, OOS)预测

function [] = tryrff_v2_function_for_each_sim(gamma, trnwin, iSim, stdize)

%**************************************************************************
% The function computes OOS performance with one random seed. 
% Parameters:
% gamma: gamma in Random Fourier Features
% trnwin: training window
% iSim: random seed for this simulation
% stdize: Standardization. stdize = 1 means True
%**************************************************************************

tic
nSim = 1; % total number of simulations run in this function

%**************************************************************************
% Choices
%**************************************************************************
% max number of Random Fourier Features (RFFs)
maxP    = 12000; 

% the grid of RFFs number
Plist   = [2 5:floor(trnwin/10):(trnwin-5) (trnwin-4):2:(trnwin+4) (trnwin+5):floor(trnwin/2):30*trnwin (31*trnwin):(10*trnwin):(maxP-1) maxP];

% training frequency
trainfrq= 1;

% shrinkage parameters lambda (z)
log_lamlist = [-3:1:3];
lamlist = 10.^(log_lamlist);

% save the result
saveon  = 1;

% Demeaning = False
demean  = 0;

% length of shrinkage parameters
nL      = length(lamlist);

% length of RFFs number grid
nP      = length(Plist);

% saving string
para_str = strcat('maxP-', 
;