matlab_通过稀疏化的klms和klms算法实现了信道均衡问题

function [expansionCoefficient,weightVector,biasTerm,learningCurve,dictionaryIndex,netSizeDiagram] = ... sparseKLMS1(trainInput,trainTarget,testInput,testTarget,typeKernel,paramKernel,... stepSizeFeatureVector,stepSizeWeightVector,stepSizeBias,toleranceDistance,tolerancePredictError,flagLearningCurve) %Function sparseKLMS1: kernel least mean square with novel criteria %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %Input: %trainInput: input signal inputDimension*trainSize, inputDimension is the input dimension and trainSize is the number of % training data %trainTarget: desired signal for training trainSize*1 %testInput: testing input, inputDimension*testSize, testSize is the number of the test data %testTarget: desired signal for testing testSize*1 %typeKernel: 'Gauss', 'Poly' %paramKernel: h (kernel size) for Gauss and p (order) for poly %stepSizeFeatureVector: learning rate for kernel part %stepSizeWeightVector: learning rate for linear part, set to zero to disable %stepSizeBias: learning rate for bias term, set to zero to disable %flagLearningCurve: control if calculating the learning curve %toleranceDistance: tolerance for the closeness of the new data to the dictionary %tolerancePredictError: tolerance for the apriori error %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %Output: %expansionCoefficient: onsisting of coefficients of the kernel % expansion with the stepSizeFeatureVector %weightVector: the linear coefficients %biasTerm: the bias term %learningCurve: trainSize*1 used for learning curve %dictionaryIndex: index of bases used in the kernel expansion in % the training set % memeory initialization trainSize = length(trainTarget); testSize = length(testTarget); if flagLearningCurve learningCurve = zeros(trainSize,1); learningCurve(1) = mean(testTarget.^2); else learningCurve =[]; end netSizeDiagram = zeros(trainSize,1); % n=1 init predictionError = trainTarget(1); expansionCoefficient = stepSizeFeatureVector*predictionError; weightVector = stepSizeWeightVector*predictionError*trainInput(:,1); biasTerm = stepSizeBias*predictionError; % dictionary dictionaryIndex = 1; dictSize = 1; netSizeDiagram(1) = 1; toleranceDistance = toleranceDistance^2; % start for n=2:trainSize % training % comparing the distance between trainInput(:,n) and the dictionary distance2dictionary = min(sum((trainInput(:,n)*ones(1,dictSize) - trainInput(:,dictionaryIndex)).^2)); if (distance2dictionary < toleranceDistance) if flagLearningCurve, learningCurve(n) = learningCurve(n-1); end netSizeDiagram(n) = netSizeDiagram(n-1); continue; end networkOutput = expansionCoefficient*ker_eval(trainInput(:,n),trainInput(:,dictionaryIndex),typeKernel,paramKernel) + weightVector'*trainInput(:,n) + biasTerm; predictionError = trainTarget(n) - networkOutput; if (abs(predictionError) < tolerancePredictError) if flagLearningCurve==1, learningCurve(n) = learningCurve(n-1); end netSizeDiagram(n) = netSizeDiagram(n-1); continue; end % updating dictSize = dictSize + 1; dictionaryIndex(dictSize) = n; expansionCoefficient(dictSize) = stepSizeFeatureVector*predictionError; netSizeDiagram(n) = netSizeDiagram(n-1) + 1; weightVector = weightVector + stepSizeWeightVector*predictionError*trainInput(:,n); biasTerm = biasTerm + stepSizeBias*predictionError; if flagLearningCurve == 1 % testing y_te = zeros(testSize,1); for jj = 1:testSize %ii = 1:dictSize; y_te(jj) = expansionCoefficient*ker_eval(testInput(:,jj),trainInput(:,dictionaryIndex),typeKernel,paramKernel) + weightVector'*testInput(:,jj) + biasTerm; end err = testTarget - y_te; learningCurve(n) = mean(err.^2); end end return