A Prediction Model for Synthetic Time Series Meta Data Fusion

In the real world, Data availability and the accessibility of a large amount of data from various sources is a major issue due to security and authentication aspects as significant obstacles. A large amount of data is required to make accurate decisions. At Present, Researchers and Data Scientists are facing challenges in reliable prediction because of less data availability. Various models are available in the literature for generating the synthetic data. The synthetic data-based model does not provide reliable results. In this paper, we have proposed a novel model for prediction by synthetic time-series data fusion which uses the concept of fusion of the original data and the synthetic data to generate the Meta Data as augmented data. It improve the model learning capacity and the prediction results. The illustration of the proposed model has been done through experiments. The results obtained from experiments show the Metadata outperforms with more accurate, reliable, and effective prediction results than the synthetic data and original data.