Abstract:
Weather forecasting is a vital component that has a tremendous impact on a modern-
day human’s day-to-day life. The modern-day human relies on the forecasted weather
data to plan their tasks. Therefore, the effectiveness and the reliability of weather
forecasting is crucial. The modern data driven world utilizes temporal weather datasets
and forecasting models for the weather forecasting process. The reliability and the
accuracy of the data generated by the forecasting models has a direct dependency on
the quality of the dataset fed into the model. The inclusion of noise in a temporal
dataset can cause the forecast model to become over complicated and inaccurate.
Hence, there is a need for a reliable and an effective mechanism to improve the quality
of temporal weather datasets.
In this research project, the author focuses on proposing a reliable and a simple
mechanism that can cleanse a temporal weather dataset ultimately enhancing its
weather forecasting performance. The proposed solution is composed with a series of
data processing techniques. The performance and the effectiveness of the implemented
system was thoroughly tested via a series of test criteria.
