Poly loss框架

PolyLoss框架是一个将损失函数看作多项式函数的线性组合的简单框架，它通过泰勒展开逼近函数。该框架允许根据目标任务和数据集来调整多项式基的重要性，并且可以将Cross-entropy loss和Focal loss作为其特殊情况。PolyLoss在论文《PolyLoss: A Polynomial Expansion Perspective of Classification Loss Functions》中被提出，并进行了广泛的实验评估。结果显示，PolyLoss在不同的任务、模型和数据集上持续提高了各方面的性能。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* *2* *3* [ICLR 2022 | 超越Focal Loss！PolyLoss用1行代码+1个超参完成超车！](https://blog.csdn.net/amusi1994/article/details/124580967)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 100%"]
[ .reference_list ]