中文摘要：

在信息技术快速发展的今天，数据形式的多样性使得对问题和现象的研究不再局限于单纯利用截面数据或时间序列数据进行分析，而是从多角度、多维度等方面对这些自然问题、经济问题和社会问题进行深入的分析和探讨。本文所研究的是一种由截面数据和时间序列数据共同组成的具有立体结构的三维数组，即时序立体数据，并以定义的点积和常数型均值为基础，试图从理论上推导出时序立体数据的回归模型。同时，以我国现代服务业的实际数据进行相应的实证分析。

英文摘要：

Along with the rapid development of information and technology, the diversity of the data forms have become the common problem confronted by data collectors and researchers. Researches on problems about nature, economy and society, are not only analyzed and discussed not only by simply using the cross-sectional data and time series data, but also in multi-angles and multi-dimensions. The analysis object in this paper is Cubic Time Series Data（CTSD）, which is threedimensional data composed by cross-sectibnal data and time ,series data. This paper tries to theoretically derive regression model of CTSD based on dot produet and constant mean, and uses data from modern service industries of our country to make empirical analysis.