KNN 手写数字识别

yiyun

发布于 2022-04-01 13:53:30

1.5K0

发布于 2022-04-01 13:53:30

文章被收录于专栏：yiyun 的专栏

图片：32*32像素黑白图像

编码

一个 3232 二进制图像矩阵转为 1 1024 的向量

# 32*32 图像矩阵 -> 1*1024 向量
def img2vector(filename):
    returnVect = zeros((1, 1024))
    fr = open(filename)
    for i in range(32):
        lineStr = fr.readline()
        for j in range(32):
            # [0, 32*i+j] 最后只有一行，遍历整个矩阵，压缩为一行即一个向量
            returnVect[0, 32*i+j] = int(lineStr[j])

    return returnVect

分类器

def classify0(inX, dataSet, labels, k):
    dataSetSize = dataSet.shape[0]
    # （以下三行）距离计算
    diffMat = tile(inX, (dataSetSize, 1)) - dataSet
    sqDiffMat = diffMat ** 2
    sqDistances = sqDiffMat.sum(axis=1)
    distances = sqDistances ** 0.5
    sortedDistIndicies = distances.argsort()

    classCount = {}
    # （以下两行）选择距离最小的k个点
    for i in range(k):
        voteIlabel = labels[sortedDistIndicies[i]]
        classCount[voteIlabel] = classCount.get(voteIlabel, 0) + 1

    # 排序
    # TODO: 没懂
    sortedClassCount = sorted(classCount.items(), key=operator.itemgetter(1), reverse=True)

    return sortedClassCount[0][0]

手写数字识别

def handwritingClassTest():
    hwLabels = []
    # 加载训练集
    trainingFileList = listdir('trainingDigits')
    m = len(trainingFileList)
    trainingMat = np.zeros((m, 1024))
    for i in range(m):
        fileNameStr = trainingFileList[i]
        # 去掉 .txt
        fileStr = fileNameStr.split('.')[0]
        # 第一个数字为分类
        classNumStr = int(fileStr.split('_')[0])
        hwLabels.append(classNumStr)
        # 一个图像矩阵转为一个行向量
        trainingMat[i, :] = img2vector('trainingDigits/%s' % fileNameStr)
    # 测试集
    testFileList = listdir('testDigits')
    errorCount = 0.0
    mTest = len(testFileList)
    for i in range(mTest):
        fileNameStr = testFileList[i]
        # 去掉 .txt
        fileStr = fileNameStr.split('.')[0]
        # 第一个数字是类别
        classNumStr = int(fileStr.split('_')[0])
        vectorUnderTest = img2vector('testDigits/%s' % fileNameStr)
        classifierResult = classify0(vectorUnderTest, trainingMat, hwLabels, 3)
        print('the classifier came back with: %d, the real answer is: %d' % (classifierResult, classNumStr))
        if (classifierResult != classNumStr): errorCount += 1.0
    print("\n the total number of errors is: %d" % errorCount)
    print("\n the total error rate is: %f" % (errorCount / float(mTest)))