A Data Asset Valuation Model of a Hybrid Machine Learning Model of Hierarchical Clustering and Long- and Short-Term Memory Networks

研究数据资产估值在宏观层面能为数字经济发展、资源配置优化、行业规范和国家竞争力提升等方面提供支撑，对推动社会经济数字化转型意义重大。在数据资产估值过程中，会遇到诸多复杂问题，如数据异质性问题、时间序列特征挖掘问题、数据维度高和复杂性问题、缺乏通用估值标准问题和突发外部事件影响问题。层次聚类和长短期记忆网络(LSTM)混合机器学习模型将层次聚类划分异质数据成簇，LSTM挖掘各簇时间序列特征，应对高维复杂数据，结合制定估值标准，快速适应突发变化；可有效应对以上诸多复杂问题。Research on data asset valuation can support the development of the digital economy, optimize resource allocation, standardize industries, and enhance national competitiveness at a macro level, playing a significant role in promoting the digital transformation of society and the economy. In the process of data asset valuation, numerous complex issues arise, such as data heterogeneity, time series feature mining, high-dimensional and complex data, lack of universal valuation standards, and the impact of sudden external events. A hybrid machine learning model combining hierarchical clustering and long short-term memory (LSTM) clusters heterogeneous data into groups through hierarchical clustering, and LSTM mines temporal features within each cluster to handle high-dimensional and complex data. By formulating valuation standards, it quickly adapts to sudden changes, effectively addressing these various complex issues.