作为一种非常有潜力的代替MLP的模型，KAN最终获得了学术界极大的关注。在我昨天的博客里，解读了最近的热门模型KAN： 论文阅读KAN: Kolmogorov–Arnold Networks-CSDN博客 KAN的原文作者提到了很多不足。本文算是对其中两个现有不足的回应，也就是:1)KAN不仅只能用于特定结构和深度，2）KAN不仅能用于小规模AI+Science任务，还可以用于更大规模或更复杂

这是一篇关于object proposal的综述，截至本文转载日期，已中了PAMI2016，目前引用次数为74。 论文的项目地址： https://www.mpi-inf.mpg.de/departments/computer-vision-and-multimodal-computing/research/object-recognition-and-scene-understanding/

题目链接：10497 - Sweet Child Makes Trouble 题意：n个物品，原来物品属于一个地方，现在要把物品重新放回去，问能放几种使得每个物品都与原来位置不同 思路：递推，一开始随便搞了个二维状态，dp[i][j]表示i个物品，有j个位置不同，那么dp[n][n]就是答案，递推式为： dp[i][j] = 1 (j == 0) dp[i] [j]

After returning from the army Makes received a gift — an array a consisting of n positive integer numbers. He hadn’t been solving problems for a long time, so he became interested to answer a particul

After returning from the army Makes received a gift — an array a consisting of npositive integer numbers. He hadn't been solving problems for a long time, so he became interested to answer a partic