-
Notifications
You must be signed in to change notification settings - Fork 9
/
SetParamsDemo2.m
20 lines (19 loc) · 1.03 KB
/
SetParamsDemo2.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
% Jun-Yan Zhu (junyanz at eecs dot berkeley dot edu)
% University of California, Berkeley
% Set the parameters for demoMCILBoost
function PARAMS = SetParamsDemo2()
PARAMS.mode = 'cross-validate'; % train, test, cross-validate
PARAMS.dataFile = fullfile('data', 'sival.data');
PARAMS.modelFile = 'sival.expr'; % in train/test model, it's the model name in 'cross-validate' model, it's the name of the folder that stores all the data
PARAMS.resultFile = 'sival.result'; % result file
CLF.verbose = 0; % 0: none; 1: some; 2: more
CLF.softmaxType = 0; % 0: Gm, 1:Lse
CLF.numWeakClfs = 150; % number of weak classifiers
CLF.numCls = 1; % number of classes in the positive bags
CLF.r = 20.0; % exponent r used in GM and LSE
PARAMS.CLF = CLF; % set classifier parameters
PARAMS.isFix = true; % if use the same data partition for cross-validation
PARAMS.nFold = 10; % #folds for cross-validation
PARAMS.thres = 0.5; % threshold for prediction
end
% corel_2, corel4