generated from LufsX/Experiment-kNN
154 lines
5.1 KiB
Python
154 lines
5.1 KiB
Python
import numpy as np
|
|
import os
|
|
from sklearn.preprocessing import StandardScaler
|
|
|
|
|
|
"""
|
|
实验说明:
|
|
- 自行编写实现数据加载与kNN的手写数字分类算法。
|
|
- 函数输入: 待分类的样本input_vector, 训练数据集dataSet, 训练集对应的标签labels。
|
|
- 函数返回: 预测的分类标签。
|
|
|
|
注意事项:
|
|
- 在function.py中实现kNN分类算法。
|
|
- 可以自行导入其他库，需要在requirements.txt添加依赖，若未在requirements.txt中添加使用到的依赖，则无法自动评测，成绩为零。
|
|
- 自动评测脚本仅会加载classify和load_data函数，请勿修改函数名。
|
|
- 训练集和测试集需手动下载解压，并放入当前目录下。
|
|
|
|
测评环境:
|
|
- CPU: Intel Xeon(R) CPU E3 - 1231 v3 @ 3.40GHz
|
|
- OS: Ubuntu 22.04.4 LTS
|
|
- Python: 3.10.12
|
|
- numpy: 2.1.1
|
|
"""
|
|
|
|
|
|
"""
|
|
请在这里填写学号和姓名，不填写即成绩无效。
|
|
"""
|
|
id = "202210423117"
|
|
name = "程悦"
|
|
|
|
|
|
def classify(input_vector, dataSet, labels, k = 2):
|
|
"""
|
|
KNN分类算法。
|
|
|
|
Args:
|
|
- input_vector (numpy array): 输入待分类的样本。
|
|
- dataSet (numpy array): 训练数据集。
|
|
- labels (list or numpy array): 训练集对应的标签。
|
|
- k (int): kNN算法参数，请**自行设置默认值**，评测时仅会使用其默认值，不修改默认值将**无法通过**评测，**成绩为零**。
|
|
|
|
Returns:
|
|
- int: 预测的分类标签。
|
|
"""
|
|
dataSet_size = dataSet.shape[0]
|
|
# 计算马氏距离
|
|
cov_matrix = np.cov(dataSet.T)
|
|
inv_cov_matrix = np.linalg.inv(cov_matrix)
|
|
diff_mat = np.tile(input_vector, (dataSet_size, 1)) - dataSet
|
|
distances = []
|
|
for i in range(dataSet_size):
|
|
distance = np.sqrt(np.dot(np.dot(diff_mat[i], inv_cov_matrix), diff_mat[i].T))
|
|
distances.append(distance)
|
|
distances = np.array(distances)
|
|
# 对距离进行排序
|
|
sorted_indices = distances.argsort()
|
|
class_count = {}
|
|
for i in range(k):
|
|
vote_label = labels[sorted_indices[i]]
|
|
class_count[vote_label] = class_count.get(vote_label, 0) + 1
|
|
sorted_class_count = sorted(class_count.items(), key = lambda x: x[1], reverse = True)
|
|
return sorted_class_count[0][0]
|
|
|
|
|
|
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。
|
|
|
|
Returns:
|
|
- np.array: 二值图像向量。
|
|
"""
|
|
return_vector = np.zeros((1, 1024))
|
|
with open(filename) as f:
|
|
for i in range(32):
|
|
line = f.readline()
|
|
for j in range(32):
|
|
return_vector[0, 32 * i + j] = int(line[j])
|
|
return return_vector
|
|
|
|
|
|
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**。
|
|
|
|
Args:
|
|
- directory (str): 数据文件夹路径。
|
|
|
|
Returns:
|
|
- np.array: 特征数据。
|
|
- np.array: 标签数据。
|
|
"""
|
|
data_set = []
|
|
labels = []
|
|
for root, dirs, files in os.walk(directory):
|
|
for file in files:
|
|
if file.endswith('.txt'):
|
|
label = int(file.split('_')[0])
|
|
labels.append(label)
|
|
data_set.append(img_to_vector(os.path.join(root, file)))
|
|
data_set = np.array(data_set).reshape(len(data_set), - 1)
|
|
scaler = StandardScaler()
|
|
data_set = scaler.fit_transform(data_set)
|
|
return data_set, np.array(labels)
|
|
|
|
|
|
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,
|
|
)
|
|
|
|
print(f"预测的标签: {predicted_label}") |