zoukankan      html  css  js  c++  java
  • How to avoid Over-fitting using Regularization?

    http://www.mit.edu/~9.520/scribe-notes/cl7.pdf

    https://en.wikipedia.org/wiki/Bayesian_interpretation_of_kernel_regularization

    the degree to which instability and complexity of the estimator should be penalized (higher penalty for increasing value of {displaystyle lambda }lambda )

    https://www.analyticsvidhya.com/blog/2015/02/avoid-over-fitting-regularization/ 

    Regularization can be motivated as a technique to improve the generalizability of a learned model.

     https://en.wikipedia.org/wiki/Regularization_(mathematics)

    Regularization can be motivated as a technique to improve the generalizability of a learned model.

    The goal of this learning problem is to find a function that fits or predicts the outcome (label) that minimizes the expected error over all possible inputs and labels. The expected error of a function f_{n} is:

    Typically in learning problems, only a subset of input data and labels are available, measured with some noise. Therefore, the expected error is unmeasurable, and the best surrogate available is the empirical error over the N available samples:

    Without bounds on the complexity of the function space (formally, the reproducing kernel Hilbert space) available, a model will be learned that incurs zero loss on the surrogate empirical error. If measurements (e.g. of x_{i}) were made with noise, this model may suffer from overfitting and display poor expected error. Regularization introduces a penalty for exploring certain regions of the function space used to build the model, which can improve generalization.

  • 相关阅读:
    HD2058The sum problem
    采用循环链表结构求解约瑟夫问题
    java线性表学习笔记(二)
    java线性表学习笔记(一)
    HD1004Let the Balloon Rise
    HD1005Number Sequence
    用c++库函数轻松解决回文问题
    accelerated C++ 中查找url(学习笔记)
    C++ 之关联容器 map
    pytorch 迁移学习[摘自官网]
  • 原文地址:https://www.cnblogs.com/rsapaper/p/7602199.html
Copyright © 2011-2022 走看看