ConceptEngineering
Newton's method
An iterative numerical method for finding successively better approximations to the roots of a real-valued function using tangent lines.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Infeasible start Newton method3Quadratic convergence3Newton step with equality constraints2Newton's method for division2Truncated Newton method2Analytic center cutting plane method1Approximate or Inexact Newton methods1Barrier method1Barrier method complexity1Basin of attraction1CG initialization1Commutative diagram1
Lecturers
8Each in their own terms, from the easiest level.
Ana Bell
A way to find the roots of a polynomial numerically using successive approximation.Stephen Boyd
An iterative method for minimizing a convex function by computing the Newton step from the Hessian and gradient using Cholesky factorization.Gilbert Strang
Hal Abelson & Gerald Jay Sussman
A general method for finding the roots (zeros) of functions by starting with a guess and iteratively applying a difference equation.Fei-Fei Li, Justin Johnson & Serena Yeung
A second-order optimization update rule that uses the inverse of the Hessian matrix to step directly to the minimum of a quadratic approximation.Srini Devadas
A method to find the roots of a function through successive approximation using tangents.Steven Strogatz
A numerical method for finding roots of equations that exhibits quadratic convergence.Jeremy Bernstein
A second-order optimization algorithm that approximates the loss function as a quadratic form and solves for the step that minimizes it.
Lectures that assume it
11- Convex Optimization II: Truncated Newton Method and L1-Norm Methods
- Square Roots, Newton's Method
- Complexity of Basic Arithmetic
- Advanced Topics and Applications
- Convex Optimization II: Localization and Cutting-Plane Methods
- Equality Constrained Minimization Continued
- Interior-Point Methods
- Trajectory Optimization I
- Modern Implicit Algorithms for Multiscale Kinetic Plasma Simulation
- Linear Programming and Two-Person Games
- Exploration of advanced topics in deep reinforcement learning
Connections
35The lectures that define it or assume it, and the concepts next to it.
Concepts10
Lectures25
- Higher-Order Procedures and Functional Abstraction
- Geometric Problems Continued
- Strings, Input/Output, and Branching in Python
- Newton's Method and Other Applications
- Training Neural Networks II
- Higher-order Procedures (SICP Lecture 2a)
Show 19 moreShow fewer
- Integer Arithmetic, Karatsuba Multiplication
- Mean Value Theorem and Inequalities
- Unconstrained Minimization
- Equality Constrained Minimization
- Feigenbaum's Renormalization Analysis of Period Doubling
- Convex Optimization II: Conjugate Gradient and Truncated Newton Methods
- Minimizing a Function Step by Step
- Scaling Rules for Optimization
- Convex Optimization II: Truncated Newton Method and L1-Norm Methods
- Square Roots, Newton's Method
- Complexity of Basic Arithmetic
- Advanced Topics and Applications
- Convex Optimization II: Localization and Cutting-Plane Methods
- Equality Constrained Minimization Continued
- Interior-Point Methods
- Trajectory Optimization I
- Modern Implicit Algorithms for Multiscale Kinetic Plasma Simulation
- Linear Programming and Two-Person Games
- Exploration of advanced topics in deep reinforcement learning