{"id":1035884,"date":"2024-12-31T12:02:31","date_gmt":"2024-12-31T04:02:31","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1035884.html"},"modified":"2024-12-31T12:02:33","modified_gmt":"2024-12-31T04:02:33","slug":"python%e6%95%b0%e6%8d%ae%e6%97%a0%e6%a0%87%e7%ad%be-%e5%a6%82%e4%bd%95%e5%8a%a0%e4%b8%8a%e6%a0%87%e7%ad%be","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1035884.html","title":{"rendered":"python\u6570\u636e\u65e0\u6807\u7b7e \u5982\u4f55\u52a0\u4e0a\u6807\u7b7e"},"content":{"rendered":"

\"python\u6570\u636e\u65e0\u6807\u7b7e<\/p>\n

\u5728Python\u4e2d\u5904\u7406\u65e0\u6807\u7b7e\u6570\u636e\u65f6\uff0c\u53ef\u4ee5\u901a\u8fc7\u591a\u79cd\u65b9\u6cd5\u7ed9\u6570\u636e\u52a0\u4e0a\u6807\u7b7e\uff0c\u5305\u62ec\u624b\u52a8\u6807\u7b7e\u3001\u81ea\u52a8\u805a\u7c7b\u3001\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b\u7b49\u3002<\/strong>\u4e0b\u9762\u662f\u8be6\u7ec6\u4ecb\u7ecd\u6bcf\u79cd\u65b9\u6cd5\u7684\u6b65\u9aa4\u548c\u5b9e\u4f8b\u3002<\/p>\n<\/p>\n

\u4e00\u3001\u624b\u52a8\u6807\u7b7e<\/h3>\n<\/p>\n

\u624b\u52a8\u6807\u7b7e\u662f\u6307\u4eba\u5de5\u6839\u636e\u6570\u636e\u7684\u5185\u5bb9\u6216\u7279\u5f81\uff0c\u624b\u52a8\u4e3a\u6bcf\u6761\u6570\u636e\u6253\u6807\u7b7e\u3002\u8fd9\u79cd\u65b9\u6cd5\u6700\u9002\u5408\u6570\u636e\u91cf\u8f83\u5c0f\u4e14\u9700\u8981\u9ad8\u7cbe\u5ea6\u6807\u7b7e\u7684\u60c5\u51b5\u3002<\/p>\n<\/p>\n

1\u3001\u6570\u636e\u9884\u5904\u7406<\/h4>\n<\/p>\n

\u5728\u8fdb\u884c\u624b\u52a8\u6807\u7b7e\u4e4b\u524d\uff0c\u9996\u5148\u9700\u8981\u5bf9\u6570\u636e\u8fdb\u884c\u9884\u5904\u7406\uff0c\u5305\u62ec\u6570\u636e\u6e05\u6d17\u3001\u683c\u5f0f\u8f6c\u6362\u7b49\u3002<\/p>\n<\/p>\n

import pandas as pd<\/p>\n
\u8bfb\u53d6\u6570\u636e<\/strong><\/h2>\n
data = pd.read_csv('your_dataset.csv')<\/p>\n
\u6570\u636e\u9884\u5904\u7406\uff08\u793a\u4f8b\uff09<\/strong><\/h2>\n
\u53bb\u9664\u7f3a\u5931\u503c<\/strong><\/h2>\n
data.dropna(inplace=True)<\/p>\n
\u6570\u636e\u683c\u5f0f\u8f6c\u6362<\/strong><\/h2>\n
data['column_name'] = data['column_name'].astype(str)<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u624b\u52a8\u6807\u7b7e<\/h4>\n<\/p>\n
\u6839\u636e\u6570\u636e\u7684\u5185\u5bb9\u6216\u7279\u5f81\uff0c\u624b\u52a8\u4e3a\u6bcf\u6761\u6570\u636e\u6253\u6807\u7b7e\u3002\u53ef\u4ee5\u4f7f\u7528Python\u7684\u8f93\u5165\u8f93\u51fa\u51fd\u6570\u5b9e\u73b0\u3002<\/p>\n<\/p>\n
# \u4e3a\u6bcf\u6761\u6570\u636e\u624b\u52a8\u6253\u6807\u7b7e<\/p>\n
labels = []<\/p>\n
for index, row in data.iterrows():<\/p>\n
    print(row['column_name'])<\/p>\n
    label = input("\u8bf7\u8f93\u5165\u8be5\u6570\u636e\u7684\u6807\u7b7e\uff1a")<\/p>\n
    labels.append(label)<\/p>\n
\u5c06\u6807\u7b7e\u6dfb\u52a0\u5230\u6570\u636e\u96c6\u4e2d<\/strong><\/h2>\n
data['label'] = labels<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e8c\u3001\u81ea\u52a8\u805a\u7c7b<\/h3>\n<\/p>\n
\u81ea\u52a8\u805a\u7c7b\u662f\u6307\u4f7f\u7528\u805a\u7c7b\u7b97\u6cd5\uff08\u5982K-means\u3001DBSCAN\u7b49\uff09\u5bf9\u6570\u636e\u8fdb\u884c\u805a\u7c7b\uff0c\u7136\u540e\u5c06\u6bcf\u4e2a\u805a\u7c7b\u7684\u7c7b\u522b\u4f5c\u4e3a\u6807\u7b7e\u3002\u8fd9\u79cd\u65b9\u6cd5\u9002\u5408\u6570\u636e\u91cf\u8f83\u5927\u4e14\u6807\u7b7e\u4e0d\u660e\u786e\u7684\u60c5\u51b5\u3002<\/p>\n<\/p>\n
1\u3001\u6570\u636e\u9884\u5904\u7406<\/h4>\n<\/p>\n
\u4e0e\u624b\u52a8\u6807\u7b7e\u7c7b\u4f3c\uff0c\u9996\u5148\u9700\u8981\u5bf9\u6570\u636e\u8fdb\u884c\u9884\u5904\u7406\u3002<\/p>\n<\/p>\n
from sklearn.preprocessing import StandardScaler<\/p>\n
\u6570\u636e\u9884\u5904\u7406\uff08\u793a\u4f8b\uff09<\/strong><\/h2>\n
scaler = StandardScaler()<\/p>\n
scaled_data = scaler.fit_transform(data)<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001K-means\u805a\u7c7b<\/h4>\n<\/p>\n
K-means\u662f\u4e00\u79cd\u5e38\u7528\u7684\u805a\u7c7b\u7b97\u6cd5\uff0c\u53ef\u4ee5\u5c06\u6570\u636e\u5206\u4e3aK\u4e2a\u805a\u7c7b\u3002<\/p>\n<\/p>\n
from sklearn.cluster import KMeans<\/p>\n
\u8bbe\u7f6e\u805a\u7c7b\u6570\u76ee<\/strong><\/h2>\n
kmeans = KMeans(n_clusters=3)<\/p>\n
\u8fdb\u884c\u805a\u7c7b<\/strong><\/h2>\n
kmeans.fit(scaled_data)<\/p>\n
\u83b7\u53d6\u805a\u7c7b\u6807\u7b7e<\/strong><\/h2>\n
labels = kmeans.labels_<\/p>\n
\u5c06\u805a\u7c7b\u6807\u7b7e\u6dfb\u52a0\u5230\u6570\u636e\u96c6\u4e2d<\/strong><\/h2>\n
data['label'] = labels<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001DBSCAN\u805a\u7c7b<\/h4>\n<\/p>\n
DBSCAN\u662f\u4e00\u79cd\u57fa\u4e8e\u5bc6\u5ea6\u7684\u805a\u7c7b\u7b97\u6cd5\uff0c\u9002\u5408\u5904\u7406\u566a\u58f0\u6570\u636e\u3002<\/p>\n<\/p>\n
from sklearn.cluster import DBSCAN<\/p>\n
\u8bbe\u7f6e\u53c2\u6570<\/strong><\/h2>\n
dbscan = DBSCAN(eps=0.5, min_samples=5)<\/p>\n
\u8fdb\u884c\u805a\u7c7b<\/strong><\/h2>\n
dbscan.fit(scaled_data)<\/p>\n
\u83b7\u53d6\u805a\u7c7b\u6807\u7b7e<\/strong><\/h2>\n
labels = dbscan.labels_<\/p>\n
\u5c06\u805a\u7c7b\u6807\u7b7e\u6dfb\u52a0\u5230\u6570\u636e\u96c6\u4e2d<\/strong><\/h2>\n
data['label'] = labels<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b<\/h3>\n<\/p>\n
\u9884\u8bad\u7ec3\u6a21\u578b\u662f\u6307\u4f7f\u7528\u5df2\u7ecf\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u5bf9\u6570\u636e\u8fdb\u884c\u5206\u7c7b\u6216\u6253\u6807\u7b7e\u3002\u5e38\u7528\u7684\u9884\u8bad\u7ec3\u6a21\u578b\u5305\u62ecBERT\u3001GPT\u7b49\u3002<\/p>\n<\/p>\n
1\u3001\u52a0\u8f7d\u9884\u8bad\u7ec3\u6a21\u578b<\/h4>\n<\/p>\n
\u4ee5BERT\u4e3a\u4f8b\uff0c\u9996\u5148\u9700\u8981\u52a0\u8f7d\u9884\u8bad\u7ec3\u6a21\u578b\u3002<\/p>\n<\/p>\n
from transformers import BertTokenizer, BertForSequenceClassification<\/p>\n
import torch<\/p>\n
\u52a0\u8f7d\u9884\u8bad\u7ec3\u6a21\u578b\u548c\u5206\u8bcd\u5668<\/strong><\/h2>\n
model = BertForSequenceClassification.from_pretrAI<\/a>ned('bert-base-uncased', num_labels=2)<\/p>\n
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u6570\u636e\u9884\u5904\u7406<\/h4>\n<\/p>\n
\u5c06\u6570\u636e\u8f6c\u6362\u4e3a\u6a21\u578b\u53ef\u4ee5\u63a5\u53d7\u7684\u683c\u5f0f\u3002<\/p>\n<\/p>\n
# \u6570\u636e\u9884\u5904\u7406\uff08\u793a\u4f8b\uff09<\/p>\n
inputs = tokenizer(data['text_column'].tolist(), return_tensors='pt', padding=True, truncation=True)<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u6a21\u578b\u9884\u6d4b<\/h4>\n<\/p>\n
\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b\u8fdb\u884c\u9884\u6d4b\uff0c\u83b7\u53d6\u6807\u7b7e\u3002<\/p>\n<\/p>\n
# \u6a21\u578b\u9884\u6d4b<\/p>\n
with torch.no_grad():<\/p>\n
    outputs = model(inputs)<\/p>\n
    predictions = torch.argmax(outputs.logits, dim=1)<\/p>\n
\u5c06\u9884\u6d4b\u6807\u7b7e\u6dfb\u52a0\u5230\u6570\u636e\u96c6\u4e2d<\/strong><\/h2>\n
data['label'] = predictions.numpy()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u56db\u3001\u6570\u636e\u53ef\u89c6\u5316\u4e0e\u8bc4\u4ef7<\/h3>\n<\/p>\n
\u5728\u7ed9\u6570\u636e\u52a0\u4e0a\u6807\u7b7e\u4e4b\u540e\uff0c\u53ef\u4ee5\u901a\u8fc7\u6570\u636e\u53ef\u89c6\u5316\u548c\u8bc4\u4ef7\u6307\u6807\u6765\u9a8c\u8bc1\u6807\u7b7e\u7684\u6709\u6548\u6027\u3002<\/p>\n<\/p>\n
1\u3001\u6570\u636e\u53ef\u89c6\u5316<\/h4>\n<\/p>\n
\u4f7f\u7528\u5e38\u7528\u7684\u6570\u636e\u53ef\u89c6\u5316\u5de5\u5177\uff08\u5982Matplotlib\u3001Seaborn\u7b49\uff09\u5bf9\u6570\u636e\u8fdb\u884c\u53ef\u89c6\u5316\u3002<\/p>\n<\/p>\n
import matplotlib.pyplot as plt<\/p>\n
import seaborn as sns<\/p>\n
\u53ef\u89c6\u5316\u6807\u7b7e\u5206\u5e03<\/strong><\/h2>\n
sns.countplot(data['label'])<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u8bc4\u4ef7\u6307\u6807<\/h4>\n<\/p>\n
\u4f7f\u7528\u5e38\u7528\u7684\u8bc4\u4ef7\u6307\u6807\uff08\u5982\u51c6\u786e\u7387\u3001\u53ec\u56de\u7387\u3001F1-score\u7b49\uff09\u5bf9\u6807\u7b7e\u8fdb\u884c\u8bc4\u4ef7\u3002<\/p>\n<\/p>\n
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score<\/p>\n
\u8bc4\u4ef7\u6807\u7b7e<\/strong><\/h2>\n
accuracy = accuracy_score(true_labels, data['label'])<\/p>\n
precision = precision_score(true_labels, data['label'], average='weighted')<\/p>\n
recall = recall_score(true_labels, data['label'], average='weighted')<\/p>\n
f1 = f1_score(true_labels, data['label'], average='weighted')<\/p>\n
print(f"Accuracy: {accuracy}")<\/p>\n
print(f"Precision: {precision}")<\/p>\n
print(f"Recall: {recall}")<\/p>\n
print(f"F1-score: {f1}")<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e94\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
\u4ee5\u4e0a\u4ecb\u7ecd\u4e86\u51e0\u79cd\u5e38\u7528\u7684\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u52a0\u4e0a\u6807\u7b7e\u7684\u65b9\u6cd5\uff0c\u5305\u62ec\u624b\u52a8\u6807\u7b7e\u3001\u81ea\u52a8\u805a\u7c7b\u3001\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b\u7b49\u3002\u6bcf\u79cd\u65b9\u6cd5\u90fd\u6709\u5176\u9002\u7528\u7684\u573a\u666f\u548c\u4f18\u7f3a\u70b9\uff0c\u9700\u8981\u6839\u636e\u5177\u4f53\u60c5\u51b5\u9009\u62e9\u5408\u9002\u7684\u65b9\u6cd5\u3002\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\uff0c\u8fd8\u53ef\u4ee5\u6839\u636e\u9700\u8981\u7ed3\u5408\u4f7f\u7528\u591a\u79cd\u65b9\u6cd5\uff0c\u4ee5\u63d0\u9ad8\u6807\u7b7e\u7684\u51c6\u786e\u6027\u548c\u6709\u6548\u6027\u3002\u603b\u7684\u6765\u8bf4\uff0c\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u52a0\u4e0a\u6807\u7b7e\u662f\u4e00\u4e2a\u590d\u6742\u4e14\u9700\u8981\u4e0d\u65ad\u8c03\u6574\u548c\u4f18\u5316\u7684\u8fc7\u7a0b\uff0c\u9700\u8981\u7ed3\u5408\u6570\u636e\u7684\u5177\u4f53\u60c5\u51b5\u548c\u4e1a\u52a1\u9700\u6c42\u8fdb\u884c\u7075\u6d3b\u5e94\u7528\u3002<\/p>\n<\/p>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
 \u5982\u4f55\u4e3a\u65e0\u6807\u7b7e\u7684Python\u6570\u636e\u96c6\u6dfb\u52a0\u6807\u7b7e\uff1f<\/strong>
\u5728Python\u4e2d\uff0c\u53ef\u4ee5\u901a\u8fc7\u591a\u79cd\u65b9\u5f0f\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u96c6\u6dfb\u52a0\u6807\u7b7e\u3002\u5e38\u89c1\u7684\u65b9\u6cd5\u5305\u62ec\u4f7f\u7528\u4eba\u5de5\u6807\u6ce8\u3001\u534a\u76d1\u7763\u5b66\u4e60\u3001\u805a\u7c7b\u7b97\u6cd5\u7b49\u3002\u4eba\u5de5\u6807\u6ce8\u662f\u6700\u76f4\u63a5\u7684\u65b9\u5f0f\uff0c\u901a\u5e38\u9700\u8981\u9886\u57df\u4e13\u5bb6\u5bf9\u6570\u636e\u8fdb\u884c\u9010\u4e00\u5ba1\u6838\u3002\u534a\u76d1\u7763\u5b66\u4e60\u5219\u7ed3\u5408\u4e86\u5c11\u91cf\u5df2\u6807\u6ce8\u6570\u636e\u4e0e\u5927\u91cf\u65e0\u6807\u7b7e\u6570\u636e\uff0c\u5229\u7528\u673a\u5668\u5b66\u4e60<\/a>\u6a21\u578b\u8fdb\u884c\u81ea\u52a8\u6807\u6ce8\u3002\u805a\u7c7b\u7b97\u6cd5\u5982K-means\u53ef\u4ee5\u5e2e\u52a9\u8bc6\u522b\u6570\u636e\u4e2d\u7684\u81ea\u7136\u5206\u7ec4\uff0c\u4e3a\u6bcf\u4e2a\u7ec4\u5206\u914d\u6807\u7b7e\u3002<\/p>\n
\u662f\u5426\u6709\u5de5\u5177\u53ef\u4ee5\u5e2e\u52a9\u6211\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u6dfb\u52a0\u6807\u7b7e\uff1f<\/strong>
\u6709\u5f88\u591a\u5de5\u5177\u548c\u5e93\u53ef\u4ee5\u8f85\u52a9\u60a8\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u6dfb\u52a0\u6807\u7b7e\u3002\u4f8b\u5982\uff0cPython\u4e2d\u7684Labelbox\u3001Prodigy\u548cDataloop\u7b49\u5e73\u53f0\u90fd\u63d0\u4f9b\u4e86\u53cb\u597d\u7684\u7528\u6237\u754c\u9762\uff0c\u652f\u6301\u5feb\u901f\u6807\u6ce8\u3002\u540c\u65f6\uff0c\u4f7f\u7528\u673a\u5668\u5b66\u4e60\u5e93\u5982scikit-learn\u548cTensorFlow\uff0c\u53ef\u4ee5\u5b9e\u73b0\u6a21\u578b\u8bad\u7ec3\u4e0e\u9884\u6d4b\uff0c\u5e2e\u52a9\u81ea\u52a8\u751f\u6210\u6807\u7b7e\u3002<\/p>\n
\u5728\u4e3a\u65e0\u6807\u7b7e\u6570\u636e\u6dfb\u52a0\u6807\u7b7e\u65f6\u5e94\u6ce8\u610f\u54ea\u4e9b\u95ee\u9898\uff1f<\/strong>
\u5728\u6dfb\u52a0\u6807\u7b7e\u65f6\uff0c\u786e\u4fdd\u6807\u7b7e\u7684\u4e00\u81f4\u6027\u548c\u51c6\u786e\u6027\u975e\u5e38\u91cd\u8981\u3002\u4f7f\u7528\u6e05\u6670\u7684\u6807\u7b7e\u6807\u51c6\u548c\u5b9a\u4e49\uff0c\u907f\u514d\u6a21\u7cca\u4e0d\u6e05\u7684\u6807\u7b7e\u3002\u6b64\u5916\uff0c\u8003\u8651\u5230\u6570\u636e\u7684\u591a\u6837\u6027\u548c\u590d\u6742\u6027\uff0c\u53ef\u80fd\u9700\u8981\u591a\u6b21\u8fed\u4ee3\u548c\u9a8c\u8bc1\uff0c\u4ee5\u4fdd\u8bc1\u6807\u7b7e\u7684\u6709\u6548\u6027\u3002\u540c\u65f6\uff0c\u4fdd\u6301\u6570\u636e\u9690\u79c1\u548c\u5408\u89c4\u6027\u4e5f\u662f\u4e0d\u53ef\u5ffd\u89c6\u7684\u65b9\u9762\uff0c\u7279\u522b\u662f\u5728\u5904\u7406\u654f\u611f\u4fe1\u606f\u65f6\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u5728Python\u4e2d\u5904\u7406\u65e0\u6807\u7b7e\u6570\u636e\u65f6\uff0c\u53ef\u4ee5\u901a\u8fc7\u591a\u79cd\u65b9\u6cd5\u7ed9\u6570\u636e\u52a0\u4e0a\u6807\u7b7e\uff0c\u5305\u62ec\u624b\u52a8\u6807\u7b7e\u3001\u81ea\u52a8\u805a\u7c7b\u3001\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b\u7b49\u3002\u4e0b […]","protected":false},"author":3,"featured_media":1035889,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1035884"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=1035884"}],"version-history":[{"count":"1","href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1035884\/revisions"}],"predecessor-version":[{"id":1035890,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1035884\/revisions\/1035890"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/1035889"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=1035884"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=1035884"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=1035884"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}