代价函数

淩亂°似流年 2022-07-15 07:43 255阅读 0赞

[Keras笔记 -- objective][Keras_ -- objective]

[Caffe中的各种loss函数适合哪些问题？][Caffe_loss]

[神经网络Loss损失函数总结][Loss]

## cross entropy ##

[交叉熵代价函数][Link 1]

[交叉熵损失函数][Link 2]

[两种交叉熵损失函数的异同][Link 3]

[Multinomial Logistic Loss and Cross Entropy Loss are the same][]

[caffe SigmoidCrossEntropyLossLayer 理论代码学习][caffe SigmoidCrossEntropyLossLayer]

## softmax ##

[UFDL-softmax loss][]

[caffe中sigmoid cross entorpy loss的计算][caffe_sigmoid cross entorpy loss]

[caffe层解读系列-softmax\_loss][caffe_-softmax_loss]

[softmax代价函数的导数计算][softmax]

[Softmax vs. Softmax-Loss: Numerical Stability][Softmax vs. Softmax-Loss_ Numerical Stability]

[caffe中带权重的softmaxloss实现（一）][caffe_softmaxloss]

https://github.com/shicai/Weighted\_Softmax\_Loss

[caffe源码分析:softmax\_layer.cpp && softmax\_loss\_layer.cpp][caffe_softmax_layer.cpp _ softmax_loss_layer.cpp]

[softmax函数][softmax 1]

## hinge loss ##

[caffe层解读系列——hinge\_loss][caffe_hinge_loss]

[caffe中HingeLossLayer层原理以及源码分析][caffe_HingeLossLayer]

[损失函数(Loss Function) -1][Loss Function_ -1]

[人脸识别：PairLoss][PairLoss]

[triplet loss 原理以及梯度推导][triplet loss]

[triplet loss layer原理][triplet loss layer]

[如何在caffe中增加layer以及caffe中triplet loss layer的实现][caffe_layer_caffe_triplet loss layer]

[Triplet Loss、Coupled Cluster Loss 探究][Triplet Loss_Coupled Cluster Loss]

[在 caffe 中添加 FaceNet 中 Triplet Loss Layer][caffe _ FaceNet _ Triplet Loss Layer]

[在caffe 中添加Scale-invariant loss][caffe _Scale-invariant loss]

## center loss ##

[ECCV16 Center Loss及其在人脸识别中的应用][ECCV16 Center Loss]

[【Technical Review】ECCV16 Grid Loss及其在人脸检测中的应用][Technical Review_ECCV16 Grid Loss]

[CenterLoss——实战&源码][CenterLoss]

[tensorflow center loss代码注释][tensorflow center loss]

# TensorFlow实现center loss #

## contrastive loss ##

[Contrastive Loss (对比损失)][Contrastive Loss]

[contrastive loss function (papers)][contrastive loss function _papers]

[损失层及其参数设置][Link 4]

[【Caffe实践】损失函数解析][Caffe]

[Caffe中Loss Layer原理的简单梳理][Caffe_Loss Layer]

## joint loss function ##

[多任务深度学习（MultiTask Learning）][MultiTask Learning]

[一箭N雕：多任务深度学习实战][N]

https://github.com/ycszen/TensorFlowLaboratory

[tf.nn.softmax\_cross\_entropy\_with\_logits的用法][tf.nn.softmax_cross_entropy_with_logits]

# [TensorFlow四种Cross Entropy算法实现和应用][TensorFlow_Cross Entropy] #

[TensorFlow损失函数专题][TensorFlow]

[Keras_ -- objective]: http://blog.csdn.net/meanme/article/details/50813719
[Caffe_loss]: http://caffecn.cn/?/question/43
[Loss]: http://blog.csdn.net/willduan1/article/details/73694826
[Link 1]: http://blog.csdn.net/u014568921/article/details/45640569
[Link 2]: http://blog.csdn.net/qq_29133371/article/details/51855157
[Link 3]: http://blog.csdn.net/u012436149/article/details/69660214
[Multinomial Logistic Loss and Cross Entropy Loss are the same]: http://blog.csdn.net/iamzhangzhuping/article/details/51459386
[caffe SigmoidCrossEntropyLossLayer]: http://blog.csdn.net/u012235274/article/details/51361290
[UFDL-softmax loss]: http://deeplearning.stanford.edu/wiki/index.php/Softmax%E5%9B%9E%E5%BD%92
[caffe_sigmoid cross entorpy loss]: http://blog.csdn.net/iamzhangzhuping/article/details/51336400
[caffe_-softmax_loss]: http://blog.csdn.net/shuzfan/article/details/51460895
[softmax]: http://blog.csdn.net/cheese_pop/article/details/51264567
[Softmax vs. Softmax-Loss_ Numerical Stability]: http://freemind.pluskid.org/machine-learning/softmax-vs-softmax-loss-numerical-stability/
[caffe_softmaxloss]: http://blog.csdn.net/langb2014/article/details/53005482
[caffe_softmax_layer.cpp _ softmax_loss_layer.cpp]: http://blog.csdn.net/cheese_pop/article/details/51122398
[softmax 1]: http://blog.csdn.net/qq_29133371/article/details/51867975
[caffe_hinge_loss]: http://blog.csdn.net/shuzfan/article/details/53032740
[caffe_HingeLossLayer]: http://blog.csdn.net/tangwei2014/article/details/46662237
[Loss Function_ -1]: http://www.cnblogs.com/rocketfan/p/4083821.html
[PairLoss]: http://blog.csdn.net/shuzfan/article/details/51525721
[triplet loss]: http://blog.csdn.net/tangwei2014/article/details/46788025
[triplet loss layer]: http://blog.csdn.net/allyli0022/article/details/73335807
[caffe_layer_caffe_triplet loss layer]: http://blog.csdn.net/tangwei2014/article/details/46812153
[Triplet Loss_Coupled Cluster Loss]: http://blog.csdn.net/u010167269/article/details/52027378
[caffe _ FaceNet _ Triplet Loss Layer]: http://blog.csdn.net/seven_first/article/details/50272005
[caffe _Scale-invariant loss]: http://blog.csdn.net/seven_first/article/details/52914542
[ECCV16 Center Loss]: http://mp.weixin.qq.com/s?__biz=MzI1NTE4NTUwOQ==&mid=2650325602&idx=1&sn=9bc6071bf62a4ccdde1df4b4c5ae39f4&chksm=f235a568c5422c7ed7993b9d3b46bcd979d360605511284e4363a4a92f845019795569f8f3e9&scene=0#rd
[Technical Review_ECCV16 Grid Loss]: http://mp.weixin.qq.com/s/qiwJnXggRwqvHFN74-W2DQ
[CenterLoss]: https://zhuanlan.zhihu.com/p/23444100
[tensorflow center loss]: http://blog.csdn.net/liyuan123zhouhui/article/details/68927502
[Contrastive Loss]: http://blog.csdn.net/autocyz/article/details/53149760
[contrastive loss function _papers]: http://blog.csdn.net/yuanchheneducn/article/details/52497061
[Link 4]: http://blog.csdn.net/u012177034/article/details/52144325
[Caffe]: http://blog.csdn.net/chenriwei2/article/details/45291739
[Caffe_Loss Layer]: http://blog.csdn.net/liumaolincycle/article/details/50822392
[MultiTask Learning]: http://blog.csdn.net/u012938704/article/details/52703262
[N]: http://mp.weixin.qq.com/s?__biz=MzI1NTE4NTUwOQ==&mid=2650325281&idx=1&sn=97779ff0da06190d6a71d33f23e9dede&mpshare=1&scene=24&srcid=1104DKMDF2K6XkRDkFbyjAvY#rd
[tf.nn.softmax_cross_entropy_with_logits]: http://blog.csdn.net/mao_xiao_feng/article/details/53382790
[TensorFlow_Cross Entropy]: http://blog.csdn.net/appleml/article/details/53861124
[TensorFlow]: http://blog.csdn.net/marsjhao/article/details/72630147