摘要：
问题：在什么场合应当用几何平均值，而不是用算术平均值？！问题：在什么场合应当用几何平均值，而不是用算术平均值？！（7.6日注，也许题目改为在什么场合使用代数（几何，…）平均值合适？更妥当）人们在很多场合（学生成绩统计、社会经济统计和科学实验等…）进行平均值的统计计算（先求N个样本的合计值再以样本数N... 阅读全文
摘要：
阅读全文
摘要：
Training Neural Networks: Q&A with Ian Goodfellow, GoogleNeural networks require considerable time and computational firepower to train. Previously, r... 阅读全文
摘要：
阅读全文
摘要：
Calculus on Computational Graphs: BackpropagationIntroductionBackpropagation is the key algorithm that makes training deep models computationally trac... 阅读全文
摘要：
Machine Learning Trick of theDay (2): Gaussian Integral TrickToday's trick, theGaussian integral trick, is one that allows us to re-expressa (potentia... 阅读全文
摘要：
Machine Learning Trick of theDay (1): Replica Trick'Tricks' of all sorts are used throughoutmachine learning, in both research and in production setti... 阅读全文
摘要：
What Does “Neurons that Fire Together Wire Together” Mean?I’ve heard the phrase “neurons that fire together wire together” more times than I can count... 阅读全文
摘要：
Reported FeaturesHow to Tell Science Stories with MapsAugust 25, 2015Greg MillerThis map, part of Audubon’sBirds and Climate Changereport, depicts pre... 阅读全文
摘要：
模型选择的一些基本思想和方法0. 引言有监督学习是日常使用最多的建模范式，它有许多更具体的名字，比如预测模型、回归模型、分类模型或者分类器。这些名字或来源统计，或来源于机器学习。关于统计学习与机器学习的区别已经有不少讨论，不少人认为机器学习侧重于目标预测，而统计学习侧重于机制理解和建模。个人更加直观... 阅读全文