| Abstract [eng] |
Master's final project presents data analysis of logistics company operating processes. Furthermore, this project provides overview of logistics company’s customers, where the main customers’ scope is shown and in which measurement system they provide orders. Project provides predicting the future of logistics company’s sales managers work results using time series analysis methods and the best model was found. Also, all the logistics company’s cargo accounting measurements entered into a single measure - loading meters, in order to accurately predict each of the company's logistics operator workload and volume. To ascertain the dummie variable effect, which represent weekends and national holidays, two models, with and without dummie variable, was compared. The changes in workload was investigated on one day, three days and a week before national holidays. These changes show, that week before national holidays have the biggest workload. In order to predict future values these methods are used: seasonal autoregressive moving average, with a designed artificial variables reflecting public holidays and weekends, model, vector autoregressive moving average model, exponential smoothing model, naive seasonal and the last value naïve model. |
