generated from LufsX/Experiment-kNN
This commit is contained in:
+50
-80
@@ -53,23 +53,23 @@ def classify(input_vector, dataSet, labels, k=3):
|
||||
Returns:
|
||||
- int: 预测的分类标签。
|
||||
"""
|
||||
# 计算输入向量与数据集中每个向量之间的距离
|
||||
distances = [np.linalg.norm(row-input_vector) for row in dataSet]
|
||||
# 获取距离最小的k个样本的索引
|
||||
sortedDistIndices = np.argsort(distances)
|
||||
# 获取这些样本对应的标签
|
||||
classCount = {}
|
||||
for i in range(k):
|
||||
voteLabel = labels[sortedDistIndices[i]]
|
||||
# 统计每个类别的出现次数
|
||||
classCount[voteLabel] = classCount.get(voteLabel, 0) + 1
|
||||
# 返回出现次数最多的类别
|
||||
sortedClassCount = sorted(classCount.items(), key=lambda x: x[1], reverse=True)
|
||||
return sortedClassCount[0][0]
|
||||
# 检查k的值是否大于训练集样本数，如果是，则抛出异常
|
||||
if k > len(labels):
|
||||
raise ValueError("k 不能大于训练集样本数")
|
||||
|
||||
# 计算输入向量与数据集中每个样本的欧几里得距离
|
||||
distances = np.linalg.norm(dataSet - input_vector, axis=1)
|
||||
# 获取距离最近的k个样本的索引
|
||||
sorted_indices = distances.argsort()
|
||||
# 获取这k个最近邻样本的标签
|
||||
nearest_labels = np.array(labels)[sorted_indices[:k]]
|
||||
# 统计这k个样本中出现次数最多的标签，并返回该标签作为预测结果
|
||||
predicted_label = np.bincount(nearest_labels).argmax()
|
||||
return predicted_label
|
||||
|
||||
def load_dataset(directory):
|
||||
"""
|
||||
从文件夹中加载数据，若不自行正确实现，则无法通过评测，**不记录成绩**。
|
||||
从文件夹中加载数据，若不自行正确实现，则无法通过评测，不记录成绩。
|
||||
|
||||
Args:
|
||||
- directory (str): 数据文件夹路径。
|
||||
@@ -78,25 +78,23 @@ def load_dataset(directory):
|
||||
- np.array: 特征数据。
|
||||
- np.array: 标签数据。
|
||||
"""
|
||||
fileList = os.listdir(directory)
|
||||
numberOfFiles = len(fileList)
|
||||
m = numberOfFiles
|
||||
n = 1024 # 假设每个图像是32x32的
|
||||
dataSet = np.zeros((m, n)) # 存储特征数据
|
||||
labels = [] # 存储标签
|
||||
for i in range(m):
|
||||
filename = fileList[i]
|
||||
filePath = os.path.join(directory, filename)
|
||||
vector = img_to_vector(filePath)
|
||||
dataSet[i, :] = vector
|
||||
# 假设文件名的格式是"label01.txt"，其中label是标签
|
||||
label = int(filename.split('.')[0][1])
|
||||
labels.append(label)
|
||||
return dataSet, np.array(labels)
|
||||
data = []
|
||||
labels = []
|
||||
# 遍历文件夹中的所有文件
|
||||
for filename in os.listdir(directory):
|
||||
if filename.endswith(".txt"):
|
||||
# 提取文件名中的标签信息
|
||||
label = int(filename.split('_')[0])
|
||||
labels.append(label)
|
||||
# 将图像文件转换为向量
|
||||
vector = img_to_vector(os.path.join(directory, filename))
|
||||
data.append(vector)
|
||||
return np.array(data), np.array(labels)
|
||||
|
||||
def img_to_vector(filename):
|
||||
"""
|
||||
将二值图像转换为向量。
|
||||
该部分应为 load_dataset 函数的辅助子函数，可单独在这里实现，也可直接集成至 load_dataset 函数中。
|
||||
|
||||
Args:
|
||||
- filename (str): 图像文件路径。
|
||||
@@ -104,57 +102,29 @@ def img_to_vector(filename):
|
||||
Returns:
|
||||
- np.array: 二值图像向量。
|
||||
"""
|
||||
returnVect = np.zeros((1, 1024)) # 假设图像是32x32的
|
||||
with open(filename, 'r') as file:
|
||||
lines = file.readlines()
|
||||
index = 0
|
||||
for line in lines:
|
||||
for i in range(len(line)-1):
|
||||
if line[i] == '1':
|
||||
returnVect[0, index] = 1
|
||||
index += 1
|
||||
return returnVect
|
||||
with open(filename, 'r') as f:
|
||||
lines = f.readlines()
|
||||
# 将32x32的字符矩阵转换为长度为1024的一维数组
|
||||
vector = np.array([int(char) for line in lines for char in line.strip()])
|
||||
return vector
|
||||
|
||||
if __name__ == "__main__":
|
||||
train_data, train_labels = load_dataset("trainingDigits")
|
||||
test_data = (
|
||||
"00000000000000110000000000000000"
|
||||
"00000000000011111000000000000000"
|
||||
"00000000000011111100000000000000"
|
||||
"00000000001111111110000000000000"
|
||||
"00000000001111111110000000000000"
|
||||
"00000000001111111100000000000000"
|
||||
"00000000011111111000000000000000"
|
||||
"00000000111111111000000000000000"
|
||||
"00000000111111110000000000000000"
|
||||
"00000000111111110000000000000000"
|
||||
"00000000111111110000000000000000"
|
||||
"00000000111111110000000000000000"
|
||||
"00000000111111100000000000000000"
|
||||
"00000001111111100000000000000000"
|
||||
"00000001111111100000000000000000"
|
||||
"00000001111111100000000000000000"
|
||||
"00000001111111111111000000000000"
|
||||
"00000001111111111111111000000000"
|
||||
"00000001111111111111111100000000"
|
||||
"00000011111111111111111110000000"
|
||||
"00000011111111111111111110000000"
|
||||
"00000001111111111111111110000000"
|
||||
"00000001111111111111111110000000"
|
||||
"00000000111111111100111111100000"
|
||||
"00000000111111111000111111100000"
|
||||
"00000000011111111001111111100000"
|
||||
"00000000011111111111111111000000"
|
||||
"00000000001111111111111111000000"
|
||||
"00000000000111111111111111000000"
|
||||
"00000000000011111111111100000000"
|
||||
"00000000000001111111110000000000"
|
||||
"00000000000000000111100000000000"
|
||||
)
|
||||
predicted_label = classify(
|
||||
np.array([int(char) for char in test_data]).reshape(1, 1024),
|
||||
train_data,
|
||||
train_labels,
|
||||
)
|
||||
# 加载训练集和测试集
|
||||
train_data, train_labels = load_dataset("D:\projects\机器学习_projects\实验\trainingDigits")
|
||||
test_data, test_labels = load_dataset("D:\projects\机器学习_projects\实验\testDigits")
|
||||
|
||||
print(f"预测的标签: {predicted_label}")
|
||||
# 记录预测的正确数量
|
||||
correct_count = 0
|
||||
|
||||
# 遍历测试集，逐个样本进行分类
|
||||
for i in range(len(test_data)):
|
||||
predicted_label = classify(test_data[i], train_data, train_labels)
|
||||
print(f"测试样本 {i+1} 的真实标签: {test_labels[i]}，预测标签: {predicted_label}")
|
||||
|
||||
# 如果预测正确，则正确数量加一
|
||||
if predicted_label == test_labels[i]:
|
||||
correct_count += 1
|
||||
|
||||
# 打印准确率
|
||||
accuracy = correct_count / len(test_data) * 100
|
||||
print(f"测试集的准确率: {accuracy:.2f}%")
|
||||
Reference in New Issue
Block a user