Nonlinear machine-learning models are increasingly used to discover causal relationships in time-series data, yet the interpretation of their outputs remains poorly understood. In particular, causal scores produced by regularized neural autoregressive models are often treated as analogues of regression coefficients, leading to misleading claims of statistical significance.
In this paper, we argue that causal relevance in nonlinear time-series models should be evaluated through forecast necessity rather than coefficient magnitude, and we present a practical evaluation procedure for doing so. We present an interpretable evaluation framework based on systematic edge ablation and forecast comparison, which tests whether a candidate causal relationship is required for accurate prediction. Using Neural Additive Vector Autoregression as a case study model, we apply this framework to a real-world case study of democratic development, modeled as a multivariate time series of panel data - democracy indicators across 139 countries. We show that relationships with similar causal scores can differ dramatically in their predictive necessity due to redundancy, temporal persistence, and regime-specific effects.
Our results demonstrate how forecast-necessity testing supports more reliable causal reasoning in applied AI systems and provides practical guidance for interpreting nonlinear time-series models in high-stakes domains.
Wavelet-Enhanced Neural ODE and Graph Attention for Interpretable Energy Forecasting
organization= School of Science, Engineering \& Technology, East Delta University , addressline = Abdullah Al Noman Road, Noman Society , city= Chattagram , postcode= 4209 , country= Bangladesh
Towards Accurate and Interpretable Time-series Forecasting: A Polynomial Learning Approach
Center for Intelligent Decision-Making（智能决策中心）；Machine Learning, School of Management, Xi'an Jiaotong University, Xi'an, China（机器学习，管理学院，西安交通大学，中国）；The 39th Research Institute of China Electronics Technology Group Corporation, Xi'an, China（中国电子科技集团第三十九研究所，中国）；Shenzhen Research Institute of Big Data, ShenZhen, China（深圳大数据研究院，中国）
Harvard University, Institute for Quantitative Social Science（哈佛大学量化社会科学研究所）；Stockholm University Demography Unit（斯德哥尔摩大学人口学单位）；Malmö University, Department of Global Political Studies（马尔默大学全球政治研究系）；University of Catania, Department of Political and Social Sciences（卡塔尼亚大学政治与社会科学系）；University of Toronto, Graduate School of Public and International Affairs（多伦多大学公共与国际事务研究生院）
Unifying Physics- and Data-Driven Modeling via Novel Causal Spatiotemporal Graph Neural Network for Interpretable Epidemic Forecasting
Frankfurt Institute for Advanced Studies（法兰克福高等研究院）；Goethe Universität Frankfurt（法兰克福大学）；Xidian-FIAS International Joint Research Center（西电-法兰克福高等研究院国际联合研究中心）；School of Science and Engineering, The Chinese University of Hong Kong（香港中文大学（深圳）理工学院）；GSI Helmholtzzentrum für Schwerionenforschung GmbH（GSI亥姆霍兹重离子研究中心）
Differentiable latent structure discovery for interpretable forecasting in clinical time series
Université Paris Cité and Université Sorbonne Paris Nord, INRIA, Inserm, HEKA（巴黎大学和巴黎-索邦大学、INRIA、Inserm、HEKA）；Department of Medical Informatics, Assistance Publique Hôpitaux de Paris (AP-HP), Georges Pompidou European Hospital（医学信息学部，巴黎公共医院（AP-HP），乔治·蓬皮杜欧洲医院）；SIERRA, Inria Paris（SIERRA，Inria巴黎）；Département d’Informatique de l’École Normale Supérieure（高等师范学院计算机系）；PSL Research University（PSL研究大学）
FDN: Interpretable Spatiotemporal Forecasting with Future Decomposition Networks
Department of Intelligent Systems Engineering, Indiana University Bloomington（印第安纳大学布卢明顿分校智能系统工程系）；Department of Computer Science and Engineering, Texas A&M University（德克萨斯大学阿姆斯特朗分校计算机科学与工程系）
iTFKAN: Interpretable Time Series Forecasting with Kolmogorov-Arnold Network
Hong Kong Polytechnic University（香港理工大学）；Northwestern Polytechnical University（西北工业大学）；Sun Yat-sen University（中山大学）；Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
MPR-Net:Multi-Scale Pattern Reproduction Guided Universality Time Series Interpretable Forecasting
Shandong University（山东大学）；Shandong Provincial Laboratory of Future Intelligence and Financial Engineering（山东省未来智能金融工程实验室）；Digital Media Technology Key Lab of Shandong Province（山东省数字媒体技术重点实验室）
A Unified Frequency Domain Decomposition Framework for Interpretable and Robust Time Series Forecasting
University of Science and Technology of China（科学技术大学）；Shanghai Black Wing Asset Management Co., Ltd.（上海黑翼资产管理有限公司）；Di-Matrix(Shanghai) Information Technology Co., Ltd.（迪矩阵（上海）信息技术有限公司）；The Chinese University of Hong Kong（香港中文大学）
Interpretable PM2.5 Forecasting for Urban Air Quality: A Comparative Study of Adaptive Time-Series Models
School of Computing, National University of Computer & Emerging Sciences (FAST)（国家计算机与新兴科学大学（FAST）计算学院）；College of Computing and Informatics, Saudi Electronic University（沙特电子大学计算与信息学院）
Platform of Bioinformatics (PBI), University Lausanne Hospital (CHUV) and Faculty of Biology and Medicine (FBM), University of Lausanne (UNIL)（生物信息平台（PBI）、日内瓦大学医院（CHUV）和生物医学学院（FBM）、日内瓦大学（UNIL））；PBI, CHUV and FBM, UNIL（PBI、CHUV和FBM、UNIL）；Department of Biomedical Engineering, University of Basel（生物医学工程系、巴塞尔大学）；Roche Diagnostics International Ltd.（罗氏诊断国际有限公司）；F. Hoffmann - La Roche Ltd., Roche Information Solutions（F. Hoffmann-La Roche有限公司、罗氏信息解决方案）
