← 返回题库
初级

实现OCR置信度三级分类

未完成
初级参考 完整示例代码供参考，建议自己理解后重新输入 
def solve():
    from pyodide.http import open_url
    from io import StringIO
    import pandas as pd
    rcp = pd.read_csv(StringIO(open_url("https://data.zuihe.com/dbd/ms-ledger/state_01/receipts.csv").read()))
    def classify(score):
        if score >= 0.9: return 'auto_accept'
        elif score >= 0.7: return 'need_confirm'
        else: return 'manual_input'
    rcp['action'] = rcp['confidence_score'].apply(classify)
    for action, grp in rcp.groupby('action'):
        print(f"  {action}: count={len(grp)} avg_conf={round(grp['confidence_score'].mean(),3)}")

示例

输入
solve()
期望输出
auto_accept: count=14 avg_conf=0.943
  manual_input: count=140 avg_conf=0.498
  need_confirm: count=46 avg_conf=0.794
Python 代码 🔒 登录后使用
🔒

登录后即可练习

注册免费账号，在浏览器中直接运行 Python 代码

登录 免费注册