Breast cancer is the most common tumor that seriously threatens the life of women. However, with imprecise measure methods, the detection results are not reliable enough, and this will bring more pain and cost to patients. Therefore, accurate identification of breast cancer is a very important issue. To tackle this problem, a composite firefly algorithm (named CoFA) is proposed, in which each firefly is attracted compositely by the best and two randomly selected fireflies. First, the composite attraction method increases the probability that the current firefly generates better solution. In addition, the two fireflies are randomly selected, whatever they are better or worse than current firefly, the population diversity can be improved. The proposed CoFA has been tested on several breast cancer datasets derived from UCI. Experimental results verified that CoFA significantly improves the recognition accuracy.

中文翻译：

---

用于萤火虫识别的复合萤火虫算法

乳腺癌是最严重威胁女性生命的最常见肿瘤。然而，采用不精确的测量方法，检测结果不够可靠，这将给患者带来更多的痛苦和成本。因此，准确识别乳腺癌是一个非常重要的问题。为了解决这个问题，提出了一种复合萤火虫算法（称为CoFA），其中，每个萤火虫被最佳和两个随机选择的萤火虫综合吸引。首先，复合吸引法增加了当前萤火虫产生更好解的可能性。另外，这两只萤火虫是随机选择的，无论它们比当前的萤火虫好还是坏，都可以改善种群多样性。拟议的CoFA已在源自UCI的多个乳腺癌数据集上进行了测试。