Forecasting discrete valued low count time series

摘要: Abstract In the past, little emphasis has been placed on producing data coherent forecasts for discrete valued processes. this paper conditional median is suggested as a general method and in contrast to conventional mean. When counts are low we suggest that of forecast be changed from forecasting future values k -step-ahead distribution. practice, usually depends unknown parameters. We modify distribution account estimation error way. The ideas exemplified by an analysis Poisson Autoregressive model wage loss claims data.