Damaged ship unsinkability classification model based on fuzzy support vector machine

When the ship is damaged after weapon attack, it is necessary for commanders to recognise its unsinkability grade quickly. Through unsinkability classification, we can know whether the ship will sink or not and its sinking probability. The unsinkability classification is a N-class pattern recognition problem. The fuzzy support vector machine (FSVM) is used to distinguish a certain unsinkability grade from other unsinkability grades firstly. Concerning the definition of fuzzy membership is critical in FSVM, the support vector data description (SVDD) is used to found fuzzy membership function. Through samples test, we found that FSVM of which fuzzy membership calculated through SVDD has better classification efficiency and precision.