Added a new notebook, interpolation_v0.3 4.ipynb, which demonstrates and compares imputation techniques for missing risk-factor data. It specifically explores Linear Interpolation against Univariate and Multivariate Kalman filters, evaluating their performance across different missing data patterns including structured outages and sporadic gaps. The notebook provides visual insights and quantitative metrics to help determine the best approach for maintaining data integrity in downstream analytics.

This update to the interpolation analysis notebook adds a robust multivariate Kalman filter approach for filling missing data. By leveraging cross-series correlations, the new implementation provides more accurate imputations compared to standard linear or univariate methods.