A Bayes‐Based Approach Against Sample Imbalance to Improving the Potential Forecasts of Gale

Sample imbalance prevents the Bayesian model from making effective potential forecasts of gale. An approach is thus proposed to modify deal with sample and verified using years (2015–2019) reanalysis radar data. The reduces by resampling based on value ranges environmental parameters application multi-layer conditional probabilities. samples that are insensitive gale excluded according for occurrence. This greatly improves occurrence hits suppresses its false alarms, leading significant improvement in forecasting skill. On basis resampling, probabilities forecast balances an equivalent magnitude. Consequently, alarms further suppressed although some reduced, resulting a higher skill