Visualization of numerical optimization algorithms

Overview

Visualization of numerical optimization algorithms

Numerical optimization is one of the core math foundations in image processing and machine learning. But I remember in the beginning of my Ph.D. years, the math behind always made me frustrated 🙁 🙁 .

During the winter vacation of 2016, I decided to make a change. I revisited some well-known optimization methods (e.g., Gradient Descent, Newton/Quasi-Newton Method, ALM, etc.), and made a series of GIF visualizations to show how these algorithms behave dynamically. Check out this repository and hope it can help you better understand these algorithms.

Gradient Descent Methods.

Fixed step size: Step size=0.5.

Fixed step size: Step size=0.1.

Fixed step size: Step size=1.

Gradient decent with the Nesterov Momentum.

Steepest Descent Method. The step size is determined by using line-search towards the gradient decent direction. The "zigzag" trajectory may cause slow convergence at ill-conditioned regions.

Conjugate Gradient Descent and Coordinate Descent Methods.

Fletcher-Reeves (FR). The FR conjugate gradient method may have very slow convergence rate if the step size is not well controlled.

Polakhe-Ribiere-Polyak (RPR). The PRP method is usually better than FR for ill-conditioned problems. Note that although it is called a "conjugate" method, the update direction (red line) is usually not vertical to the true gradient direction (black line).

Coordinate Descent. The coordinate descent method selects only one coordinate at one time for update. The well-known LibLinear package incorporates this idea to solve the linear SVM. In ill-conditioned regions, this algorithm may also face the "zigzag-step" problem.

Newton Methods.

Basic Newton Method. The black curve is the contour of the 2nd order approximation of the objective function. As the Hessian matrix at the initial point is non-positive, the optimization is not stable at very early steps.

Levenbery-Marquardt (LM) Method. LM method improves the stability of the basic Newton method by adding a small diagonal matrix to the Hessian matrix. This algorithm also can be seen as an integration of the basic Newton method and the gradient descent method.

Damped Newton Method. Damped Newton method can be viewed as a combination of the basic Newton method and the line-search based method. In spite of the fact that the Hessian matrix may be non-positive, the convergance can still be guaranteed.

Broyden Fletcher Goldfarb Shanno (BFGS). The BFGS method is the representative of quasi-Newton methods. It takes the first order gradients to approximate the Hessian matrix. In this figure, the red curve represents the true second-order information, while the black curve represents an approximated one by using BFGS.

Gaussian Newton Least Square Method (GNLS). The Gaussian-Newton least square method is a classical algorithm for solving nonlinear least squares regression problems. The essence of this algorithm is to use the first order Jacobian matrix (black curve) as an approximation of the Hessian matrix (red curve).

Random search algorithm

Genetic Algorithm (GA). GA is a classical algorithm to solve non-convex optimization problems. The key to this algorithm can be summarized as: "breeding", "mutation" and "natural selection". In this figure, the green scatters represent the descendants and the red ones represent the result of natural selection.

Simulated Annealing Algorithm (SAA). SAA is another kind of classical algorithm to solve nonconvex optimization problems. In this figure, the red curve on the right corresponds to the "temperature" and the blue curve corresponds to the objective function value. The objective function value converges with the decrease of the temperature.

Constrained Optimization Method

Gradient Projection Method (GPM). GPM is the most straight-forward way to solve a constrained optimization problem. In each interation, the gradient is projected to the feasible domain to make the current point satisfies the constraints.

Exterior-Point Penalty Method. The exterior-point penalty method is a classical way to solve constrained optimization problems. The key to this algorithm is to penalize the objective function outside the feasible domain so that to convert the original constrained problem into an unconstrained one. Note that the objectve may become ill-conditioned at the boundary of the constraints.

Inner-Point Barrier Method. The Inner-Point Barrier Method is another classical way to solve constrained optimization problems. Different from the exterior-point penalty methods where the objective is penalized outside the feasible region, the inner-point barrier method constructs a barrier function at the boundary of the feasible domain so that to prevent crossing the boundary. Similar to the exterior-point penalty method, the objectve may become ill-conditioned at the boundary of the constraints.

Lagrange Dual Ascent Method. By adding a Lagrangian multiplier, any constrained problem can be equally converted to an unconstrained max-min problem . In the Lagrange Dual Ascent Method, the variable x and the Lagrangian multiplier coefficient are alternately updated. Note that when the background color changes, the Lagrangian multiplier started to be taken into consideration during the updates.

Augmented Lagrangian Method (ALM). ALM is designed based on the Lagrange Dual Ascent Method by adding a penalty function as Augmented Lagrangian multipliers. ALM is more robust at ill-conditioned regions, e.g., at the boundary of constraints.

"keep Calm and Don't Overfit."

Owner
Zhengxia Zou
Postdoc at the University of Michigan. Research interest: computer vision and applications in remote sensing, self-driving, and video games.
Zhengxia Zou
Decision Border Visualizer for Classification Algorithms

dbv Decision Border Visualizer for Classification Algorithms Project description A python package for Machine Learning Engineers who want to visualize

Sven Eschlbeck 1 Nov 01, 2021
Main repository for Vispy

VisPy: interactive scientific visualization in Python Main website: http://vispy.org VisPy is a high-performance interactive 2D/3D data visualization

vispy 3k Jan 03, 2023
Make visual music sheets for thatskygame (graphical representations of the Sky keyboard)

sky-python-music-sheet-maker This program lets you make visual music sheets for Sky: Children of the Light. It will ask you a few questions, and does

21 Aug 26, 2022
Plot, scatter plots and histograms in the terminal using braille dots

Plot, scatter plots and histograms in the terminal using braille dots, with (almost) no dependancies. Plot with color or make complex figures - similar to a very small sibling to matplotlib. Or use t

Tammo Ippen 207 Dec 30, 2022
Massively parallel self-organizing maps: accelerate training on multicore CPUs, GPUs, and clusters

Somoclu Somoclu is a massively parallel implementation of self-organizing maps. It exploits multicore CPUs, it is able to rely on MPI for distributing

Peter Wittek 239 Nov 10, 2022
University of Missouri - Kansas City: CS451R: Capstone

CS451RC University of Missouri - Kansas City: CS451R: Capstone Installation cd git clone https://github.com/ala2q6/CS451RC.git cd CS451RC pip3 instal

Alex Arbuckle 1 Nov 17, 2021
Flame Graphs visualize profiled code

Flame Graphs visualize profiled code

Brendan Gregg 14.1k Jan 03, 2023
Visualize large time-series data in plotly

plotly_resampler enables visualizing large sequential data by adding resampling functionality to Plotly figures. In this Plotly-Resampler demo over 11

PreDiCT.IDLab 604 Dec 28, 2022
Generate knowledge graphs with interesting geometries, like lattices

Geometric Graphs Generate knowledge graphs with interesting geometries, like lattices. Works on Python 3.9+ because it uses cool new features. Get out

Charles Tapley Hoyt 5 Jan 03, 2022
In-memory Graph Database and Knowledge Graph with Natural Language Interface, compatible with Pandas

CogniPy for Pandas - In-memory Graph Database and Knowledge Graph with Natural Language Interface Whats in the box Reasoning, exploration of RDF/OWL,

Cognitum Octopus 34 Dec 13, 2022
2D maze path solver visualizer implemented with python

2D maze path solver visualizer implemented with python

SS 14 Dec 21, 2022
PolytopeSampler is a Matlab implementation of constrained Riemannian Hamiltonian Monte Carlo for sampling from high dimensional disributions on polytopes

PolytopeSampler PolytopeSampler is a Matlab implementation of constrained Riemannian Hamiltonian Monte Carlo for sampling from high dimensional disrib

9 Sep 26, 2022
Fractals plotted on MatPlotLib in Python.

About The Project Learning more about fractals through the process of visualization. Built With Matplotlib Numpy License This project is licensed unde

Akeel Ather Medina 2 Aug 30, 2022
Set of matplotlib operations that are not trivial

Matplotlib Snippets This repository contains a set of matplotlib operations that are not trivial. Histograms Histogram with bins adapted to log scale

Raphael Meudec 1 Nov 15, 2021
A package for plotting maps in R with ggplot2

Attention! Google has recently changed its API requirements, and ggmap users are now required to register with Google. From a user’s perspective, ther

David Kahle 719 Jan 04, 2023
a plottling library for python, based on D3

Hello August 2013 Hello! Maybe you're looking for a nice Python interface to build interactive, javascript based plots that look as nice as all those

Mike Dewar 1.4k Dec 28, 2022
Python Data Structures for Humans™.

Schematics Python Data Structures for Humans™. About Project documentation: https://schematics.readthedocs.io/en/latest/ Schematics is a Python librar

Schematics 2.5k Dec 28, 2022
Streamlit-template - A streamlit app template based on streamlit-option-menu

streamlit-template A streamlit app template for geospatial applications based on

Qiusheng Wu 41 Dec 10, 2022
Tweets your monthly GitHub Contributions as Wordle grid

Tweets your monthly GitHub Contributions as Wordle grid

Venu Vardhan Reddy Tekula 5 Feb 16, 2022
Calendar heatmaps from Pandas time series data

Note: See MarvinT/calmap for the maintained version of the project. That is also the version that gets published to PyPI and it has received several f

Martijn Vermaat 195 Dec 22, 2022