Strategies for Producing Reliable Trends Forecasting of COVID-19 Pandemic in Malaysia using Dynamic Mode Decomposition

Dynamic Mode Decomposition (DMD) with time delay embedding is used to predict dynamic patterns in univariate series. An important pattern that can be extracted using DMD the trend or global change a series which useful for producing reliable forecast. utilizes computationally effi cient singular value decomposition (SVD) produce low rank approximation of linear operator brings about Trend translated as modes frequencies. The evolution this frequency produces forecast In paper, we outline strategies extracting component from COVID-19 Malaysia. It discovered that, other than identifying slow varying frequencies, need also resolve stamp delay, so mean-square error reconstructed minimal. Information magnitude and phase are identify persistent remove nonstationary ones. We compare performance another SVD-based method spectrum analysis (SSA) our results highlight certain fundamental difference between these two methods. forecasts SSA tend lean towards direction maximum variance, reconstruction but detect sudden changes On hand, captures phases dominant dictates overall pattern, hence providing better prediction future dynamics