ML之回歸預測之Lasso：利用Lasso算法解決回歸(實數值評分預測)問題—優化模型【增加新(組合)屬性】

輸出結果

設計思路

核心代碼

names[-1] = "a^2"

names.append("a*b")

nrows = len(xList)

ncols = len(xList[0])

xMeans = []

xSD = []

for i in range(ncols):

   col = [xList[j][i] for j in range(nrows)]

   mean = sum(col)/nrows

   xMeans.append(mean)

   colDiff = [(xList[j][i] - mean) for j in range(nrows)]

   sumSq = sum([colDiff[i] * colDiff[i] for i in range(nrows)])

   stdDev = sqrt(sumSq/nrows)

   xSD.append(stdDev)

X = numpy.array(xList)             #Unnormalized X's

X = numpy.array(xNormalized)       #Normlized Xss

Y = numpy.array(labels)            #Unnormalized labels