To study battle-field simulation methods based on Vega,a virtual battle-field simulated by an imaginary combat happened on the sea was designed. The simulation framework in the sea battle-filed included helicopter simulation,fire simulation, collision detection and detonation, and simulation of dynamic sea surface. The method to build the simulation environments and actions to them was discussed. And the simulation experiments were conducted.,It is indicated that the simulated sea battle-field based on Vega is feasible and helpful for forces and battle-field.
Tri-training能有效利用无标记样例提高泛化能力.针对Tri-training迭代中无标记样例常被错误标记而形成训练集噪声,导致性能不稳定的缺点,文中提出ADE-Tri-training(Tri-training with Adaptive Data Editing)新算法.它不仅利用Remove Only剪辑操作对每次迭代可能产生的误标记样例识别并移除,更重要的是采用自适应策略来确定Remove Only触发与抑制的恰当时机.文中证明,PAC理论下自适应策略中一系列判别充分条件可同时确保新训练集规模迭代增大和新假设分类错误率迭代降低更多.UCI数据集上实验结果表明:ADE-Tri-training具有更好的分类泛化性能和健壮性.
