function abClassifier = buildAdaBoost(trnX, trnY, iter, tstX, tstY)

if nargin < 4

    tstX = [];

    tstY = [];

end

abClassifier = initAdaBoost(iter);

N = size(trnX, 1); % Number of training samples

sampleWeight = repmat(1/N, N, 1);

for i = 1:iter

    weakClassifier = buildStump(trnX, trnY, sampleWeight);

    abClassifier.WeakClas{i} = weakClassifier;

    abClassifier.nWC = i;

    % Compute the weight of this classifier

    abClassifier.Weight(i) = 0.5*log((1-weakClassifier.error)/weakClassifier.error);

    % Update sample weight

    label = predStump(trnX, weakClassifier);

    tmpSampleWeight = -1*abClassifier.Weight(i)*(trnY.*label); % N x 1

    tmpSampleWeight = sampleWeight.*exp(tmpSampleWeight); % N x 1

    sampleWeight = tmpSampleWeight./sum(tmpSampleWeight); % Normalized

    % Predict on training data

    [ttt, abClassifier.trnErr(i)] = predAdaBoost(abClassifier, trnX, trnY);

    % Predict on test data

    if ~isempty(tstY)

        abClassifier.hasTestData = true;

        [ttt, abClassifier.tstErr(i)] = predAdaBoost(abClassifier, tstX, tstY);

    end

    % fprintf('\tIteration %d, Training error %f\n', i, abClassifier.trnErr(i));

end

end