What is the best setup of PMM to improve imputation results within missing gap and single missing over time data? #97
Description
Hi,
thanks for your repo and I saw a post from someone in LinkedIn here and motivated to reproduce the same results on similar generated data as well as real scenario over uni-variante time-series data with missing gaps however i could not manage to find the best setup. can you comment on my experiment in provided notebook:
 |
 |
|---|---|
| Fig. 1: Comparison of imputations for missing (simple scenario shared in LinkedIn) | Fig. 2: Comparison of imputations for single missing sequence (real scenario) |
Looking forward to hearing your tips to improve the PMM results in both scenario