| Summary: | 碩士 === 國立臺灣海洋大學 === 商船學系所 === 99 === Although the shipping companies make a request that captains should obey the company’s instruction. For the sake of safety, the captain should consider whether the instruction is correct. For this reason, it is important to build a model of route planning which is different from the scenario used by shipping company.
Base on the principle of safety, the model’s independent variables are divided into two primary parts including weather conditions and ship conditions. The research has obtained the logistic regression model (LRM) through the logistic regression analysis (LRA).
According to the data output of the Logistic Regression Analysis, using continuous variables as independent variables cause poor model goodness of fit. But transfer these variables into categorical variables makes model goodness of fit obviously rising and all the variables can pass the significance test. If use true data as input data can get the same conclusion, it will be a great benefit to develops the model to practical phase. Users only need to input simple code instead of complicated detail to achieve the purpose of route estimation.
According to the concept of route choice, the research successfully build a reasonable choice model by Logistic Regression Analysis. We can infer that this model could be more representative while using practical sailing data.
|
|---|
