Hi fellow Data Science and AI members,

I've hit a roadblock on a project and could use some expert guidance. Specifically, I'm struggling with implementing an interpretable machine learning model for time series forecasting. The models I've tried lack the transparency and explainability my stakeholders require. Recurrent Neural Networks work well but their predictions are opaque. Is anyone familiar with approaches like SHAP values, LIME, or inherently interpretable models (e.g., RuleFit, Generalized Additive Models) that can provide meaningful explanations for time-based predictions? I'd greatly appreciate any resources, code snippets, or advice you could share. Thanks in advance for your help!