摘要
Abstract
To address such key challenges in analyzing influencing factors of petrochemical product prices as the high dimensionality and complexity of data,strong dynamic interdependencies,and the limited interpretability of traditional machine learning models,the paper proposes an explainable analytical framework integrating Social Network Analysis(SNA),the Relative Strength Index(RSI),and Dynamic Time Warping(DTW).It identifies the hierarchical structure of key influencing factors by constructing a correlation network of chemical product prices,obtains standardize multi-dimensional data and extract trend features using The RSI,sorts out algorithm quantifies dynamic temporal correlations combining the DTW,and finally builds an interpretable decision-making model with explicit weights by the Analytic Hierarchy Process(AHP).Compared with conventional black-box models,it not only preserves the capacity to capture temporal characteristics but also visually reveals the pathways through such core mechanisms as import/export prices,industrial chain transmission,and substitute product competition to offer a solution for price fluctuation attribution in the petrochemical industry under big data contexts,balancing analytical accuracy with interpretability.The research outcomes have been formalized into a reusable intelligent analysis template,serving as a demonstration of the value of big data mining in the energy and chemical sectors and providing new perspectives and methodologies for the analysis of petrochemical product prices.关键词
石化产品价格/社会网络分析/相对强弱指数/动态时间规整/价格分析/层次分析法/可解释性Key words
petrochemical product prices/social network analysis/relative strength index/dynamic time warping/price analysis/analytic hierarchy process/interpretability分类
管理科学
