R11 练习: 大数据技术
考纲范围
- Describe aspects of “fintech” that are directly relevant for the gathering and analyzing of financial data.
- Describe Big Data, artificial intelligence, and machine learning.
- Describe applications of Big Data and Data Science to investment management.
Q1.
Which of the following statements about fintech in investment management is least accurate?
A. Investors benefit from automated trading for more efficiency and lower transaction cost.
B. Development of fintech only affects the quantitative asset managers but not the fundamental asset managers in investment decision-making.
C. Fintech has advanced into decision-making applications based on complex machine-learning logic.
查看答案与解析
答案：B
解析：金融科技对投资管理的影响。
选项 判断 解析 A ✓ 正确，自动化交易提高效率并降低交易成本 B ✗ 不正确。金融科技不仅影响量化资产管理者，也影响基本面分析的资产管理者。例如NLP可用于分析财报、新闻等 C ✓ 正确，金融科技已发展到基于复杂机器学习逻辑的决策应用
Q2.
All the following are characteristics of Big Data except that:
A. almost no problems exist in investment analysis.
B. near real-time communication.
C. sources from traditional data and alternative data.
查看答案与解析
答案：A
解析：大数据的特征和挑战。
选项 判断 解析 A 不是特征 大数据在投资分析中仍存在许多问题和挑战（如数据质量、过拟合、隐私等），说”几乎没有问题”是不正确的 B 是特征 大数据具有近实时通信的特征（Velocity） C 是特征 大数据的来源包括传统数据和另类数据（Variety）
Q3.
Which of the following descriptions is least accurate? Machine learning:
A. seeks to extract knowledge from data without making any assumptions on the data’s underlying probability distribution.
B. splits the dataset into two distinct subsets: training dataset and validation dataset.
C. may be subject to underfitting issues because the model treats true parameters as noise.
查看答案与解析
答案：B
解析：机器学习的特征和数据集划分。
选项 判断 解析 A ✓ 正确，机器学习不需要对数据的底层概率分布做假设 B ✗ 不够准确。机器学习通常将数据集分为三个子集：训练集（training）、验证集（validation）和测试集（test），不只是两个 C ✓ 正确，欠拟合（underfitting）是指模型过于简单，将真实参数视为噪声
Q4.
Natural language processing (NLP) is appropriate to be used in:
A. identifying trends in longer timespans.
B. speech recognition and sentiment analysis.
C. incorporating traditional information to identify long-term indicators about an economic event.
查看答案与解析
答案：B
解析：NLP（自然语言处理）的应用场景。
选项 判断 解析 A ✗ 识别长期趋势更适合时间序列分析，不是NLP的主要应用 B ✓ NLP的核心应用包括语音识别和情感分析 C ✗ 整合传统信息识别长期经济指标不是NLP的典型应用