Implemented fully documented Particle Swarm Optimization algorithm (basic model with few advanced features) using Python programming language

Overview

Enhanced Particle Swarm Optimization (PSO) with Python

GitHub license GitHub issues

Implemented fully documented Particle Swarm Optimization (PSO) algorithm in Python which includes a basic model along with few advanced features such as updating inertia weight, cognitive, social learning coefficients and maximum velocity of the particle.

Dependencies

  • Numpy
  • matplotlib

Utilities

Once the installation is finished (download or cloning), go the pso folder and follow the below simple guidelines to execute PSO effectively (either write the code in command line or in a python editor).

>>> from pso import PSO

Next, a fitness function (or cost function) is required. I have included four different fitness functions for example purposes namely fitness_1, fitness_2, fitness_3, and fitness_4.

Fitness-1 (Himmelblau's Function)

Minimize: f(x) = (x2 + y - 11)2 + (x + y2 - 7)2

Optimum solution: x = 3 ; y = 2

Fitness-2 (Booth's Function)

Minimize: f(x) = (x + 2y - 7)2 + (2x + y - 5)2

Optimum solution: x = 1 ; y = 3

Fitness-3 (Beale's Function)

Minimize: f(x) = (1.5 - x - xy)2 + (2.25 - x + xy2)2 + (2.625 - x + xy3)2

Optimum solution: x = 3 ; y = 0.5

Fitness-4

Maximize: f(x) = 2xy + 2x - x2 - 2y2

Optimum solution: x = 2 ; y = 1

>>> from fitness import fitness_1, fitness_2, fitness_3, fitness_4

Now, if you want, you can provide an initial position X0 and bound value for all the particles (not mandatory) and optimize (minimize or maximize) the fitness function using PSO:

NOTE: a bool variable min=True (default value) for MINIMIZATION PROBLEM and min=False for MAXIMIZATION PROBLEM

>>> PSO(fitness=fitness_1, X0=[1,1], bound=[(-4,4),(-4,4)]).execute()

You will see the following similar output:

OPTIMUM SOLUTION
  > [3.0000078, 1.9999873]

OPTIMUM FITNESS
  > 0.0

When fitness_4 is used, observe that min=False since it is a Maximization problem.

>>> PSO(fitness=fitness_4, X0=[1,1], bound=[(-4,4),(-4,4)], min=False).execute()

You will see the following similar output:

OPTIMUM SOLUTION
  > [2.0, 1.0]

OPTIMUM FITNESS
  > 2.0

Incase you want to print the fitness value for each iteration, then set verbose=True (here Tmax=50 is the maximum iteration)

>>> PSO(fitness=fitness_2, Tmax=50, verbose=True).execute()

You will see the following similar output:

Iteration:   0  | best global fitness (cost): 18.298822
Iteration:   1  | best global fitness (cost): 1.2203953
Iteration:   2  | best global fitness (cost): 0.8178153
Iteration:   3  | best global fitness (cost): 0.5902262
Iteration:   4  | best global fitness (cost): 0.166928
Iteration:   5  | best global fitness (cost): 0.0926638
Iteration:   6  | best global fitness (cost): 0.0926638
Iteration:   7  | best global fitness (cost): 0.0114517
Iteration:   8  | best global fitness (cost): 0.0114517
Iteration:   9  | best global fitness (cost): 0.0114517
Iteration:   10 | best global fitness (cost): 0.0078867
Iteration:   11 | best global fitness (cost): 0.0078867
Iteration:   12 | best global fitness (cost): 0.0078867
Iteration:   13 | best global fitness (cost): 0.0078867
Iteration:   14 | best global fitness (cost): 0.0069544
Iteration:   15 | best global fitness (cost): 0.0063058
Iteration:   16 | best global fitness (cost): 0.0063058
Iteration:   17 | best global fitness (cost): 0.0011039
Iteration:   18 | best global fitness (cost): 0.0011039
Iteration:   19 | best global fitness (cost): 0.0011039
Iteration:   20 | best global fitness (cost): 0.0011039
Iteration:   21 | best global fitness (cost): 0.0007225
Iteration:   22 | best global fitness (cost): 0.0005875
Iteration:   23 | best global fitness (cost): 0.0001595
Iteration:   24 | best global fitness (cost): 0.0001595
Iteration:   25 | best global fitness (cost): 0.0001595
Iteration:   26 | best global fitness (cost): 0.0001595
Iteration:   27 | best global fitness (cost): 0.0001178
Iteration:   28 | best global fitness (cost): 0.0001178
Iteration:   29 | best global fitness (cost): 0.0001178
Iteration:   30 | best global fitness (cost): 0.0001178
Iteration:   31 | best global fitness (cost): 0.0001178
Iteration:   32 | best global fitness (cost): 0.0001178
Iteration:   33 | best global fitness (cost): 0.0001178
Iteration:   34 | best global fitness (cost): 0.0001178
Iteration:   35 | best global fitness (cost): 0.0001178
Iteration:   36 | best global fitness (cost): 0.0001178
Iteration:   37 | best global fitness (cost): 2.91e-05
Iteration:   38 | best global fitness (cost): 1.12e-05
Iteration:   39 | best global fitness (cost): 1.12e-05
Iteration:   40 | best global fitness (cost): 1.12e-05
Iteration:   41 | best global fitness (cost): 1.12e-05
Iteration:   42 | best global fitness (cost): 1.12e-05
Iteration:   43 | best global fitness (cost): 1.12e-05
Iteration:   44 | best global fitness (cost): 1.12e-05
Iteration:   45 | best global fitness (cost): 1.12e-05
Iteration:   46 | best global fitness (cost): 1.12e-05
Iteration:   47 | best global fitness (cost): 2.4e-06
Iteration:   48 | best global fitness (cost): 2.4e-06
Iteration:   49 | best global fitness (cost): 2.4e-06
Iteration:   50 | best global fitness (cost): 2.4e-06

OPTIMUM SOLUTION
  > [1.0004123, 2.9990281]

OPTIMUM FITNESS
  > 2.4e-06

Now, incase you want to plot the fitness value for each iteration, then set plot=True (here Tmax=50 is the maximum iteration)

>>> PSO(fitness=fitness_2, Tmax=50, plot=True).execute()

You will see the following similar output:

OPTIMUM SOLUTION
  > [1.0028365, 2.9977422]

OPTIMUM FITNESS
  > 1.45e-05

Fitness

Finally, in case you want to use the advanced features as mentioned above (say you want to update the weight inertia parameter w), simply use update_w=True and thats it. Similarly you can use update_c1=True (to update individual cognitive parameter c1), update_c2=True (to update social learning parameter c2), and update_vmax=True (to update maximum limited velocity of the particle vmax)

>>> PSO(fitness=fitness_1, update_w=True, update_c1=True).execute()

References:

[1] Almeida, Bruno & Coppo leite, Victor. (2019). Particle swarm optimization: a powerful technique for solving engineering problems. 10.5772/intechopen.89633.

[2] He, Yan & Ma, Wei & Zhang, Ji. (2016). The parameters selection of pso algorithm influencing on performance of fault diagnosis. matec web of conferences. 63. 02019. 10.1051/matecconf/20166302019.

[3] Clerc, M., and J. Kennedy. The particle swarm — explosion, stability, and convergence in a multidimensional complex space. ieee transactions on evolutionary computation 6, no. 1 (february 2002): 58–73.

[4] Y. H. Shi and R. C. Eberhart, “A modified particle swarm optimizer,” in proceedings of the ieee international conferences on evolutionary computation, pp. 69–73, anchorage, alaska, usa, may 1998.

[5] G. Sermpinis, K. Theofilatos, A. Karathanasopoulos, E. F. Georgopoulos, & C. Dunis, Forecasting foreign exchange rates with adaptive neural networks using radial-basis functions and particle swarm optimization, european journal of operational research.

[6] Particle swarm optimization (pso) visually explained (https://towardsdatascience.com/particle-swarm-optimization-visually-explained-46289eeb2e14)

[7] Rajib Kumar Bhattacharjya, Introduction to Particle Swarm Optimization (http://www.iitg.ac.in/rkbc/ce602/ce602/particle%20swarm%20algorithms.pdf)

Code repository for the paper: Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the Wild (ICCV 2021)

Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the Wild Akash Sengupta, Ignas Budvytis, Robert

Akash Sengupta 149 Dec 14, 2022
EFENet: Reference-based Video Super-Resolution with Enhanced Flow Estimation

EFENet EFENet: Reference-based Video Super-Resolution with Enhanced Flow Estimation Code is a bit messy now. I woud clean up soon. For training the EF

Yaping Zhao 19 Nov 05, 2022
AFLNet: A Greybox Fuzzer for Network Protocols

AFLNet: A Greybox Fuzzer for Network Protocols AFLNet is a greybox fuzzer for protocol implementations. Unlike existing protocol fuzzers, it takes a m

626 Jan 06, 2023
Weakly Supervised Dense Event Captioning in Videos, i.e. generating multiple sentence descriptions for a video in a weakly-supervised manner.

WSDEC This is the official repo for our NeurIPS paper Weakly Supervised Dense Event Captioning in Videos. Description Repo directories ./: global conf

Melon(Xuguang Duan) 96 Nov 01, 2022
Customizable RecSys Simulator for OpenAI Gym

gym-recsys: Customizable RecSys Simulator for OpenAI Gym Installation | How to use | Examples | Citation This package describes an OpenAI Gym interfac

Xingdong Zuo 14 Dec 08, 2022
imbalanced-DL: Deep Imbalanced Learning in Python

imbalanced-DL: Deep Imbalanced Learning in Python Overview imbalanced-DL (imported as imbalanceddl) is a Python package designed to make deep imbalanc

NTUCSIE CLLab 19 Dec 28, 2022
Official implementation of the paper 'Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution' in CVPR 2022

LDL Paper | Supplementary Material Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution Jie Liang*, Hu

150 Dec 26, 2022
PyTorch implementaton of our CVPR 2021 paper "Bridging the Visual Gap: Wide-Range Image Blending"

Bridging the Visual Gap: Wide-Range Image Blending PyTorch implementaton of our CVPR 2021 paper "Bridging the Visual Gap: Wide-Range Image Blending".

Chia-Ni Lu 69 Dec 20, 2022
Vignette is a face tracking software for characters using osu!framework.

Vignette is a face tracking software for characters using osu!framework. Unlike most solutions, Vignette is: Made with osu!framework, the game framewo

Vignette 412 Dec 28, 2022
[ACL-IJCNLP 2021] Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning

CLNER The code is for our ACL-IJCNLP 2021 paper: Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning CLNER is a

71 Dec 08, 2022
Codebase for the Summary Loop paper at ACL2020

Summary Loop This repository contains the code for ACL2020 paper: The Summary Loop: Learning to Write Abstractive Summaries Without Examples. Training

Canny Lab @ The University of California, Berkeley 44 Nov 04, 2022
a dnn ai project to classify which food people are eating on audio recordings

Deep Learning - EAT Challenge About This project is part of an AI challenge of the DeepLearning course 2021 at the University of Augsburg. The objecti

Marco Tröster 1 Oct 24, 2021
The code succinctly shows how our ensemble learning based on deep learning CNN is used for LAM-avulsion-diagnosis.

deep-learning-LAM-avulsion-diagnosis The code succinctly shows how our ensemble learning based on deep learning CNN is used for LAM-avulsion-diagnosis

1 Jan 12, 2022
PipeTransformer: Automated Elastic Pipelining for Distributed Training of Large-scale Models

PipeTransformer: Automated Elastic Pipelining for Distributed Training of Large-scale Models This repository is the official implementation of the fol

DistributedML 41 Dec 06, 2022
Geometric Vector Perceptrons --- a rotation-equivariant GNN for learning from biomolecular structure

Geometric Vector Perceptron Implementation of equivariant GVP-GNNs as described in Learning from Protein Structure with Geometric Vector Perceptrons b

Dror Lab 142 Dec 29, 2022
Python Implementation of the CoronaWarnApp (CWA) Event Registration

Python implementation of the Corona-Warn-App (CWA) Event Registration This is an implementation of the Protocol used to generate event and location QR

MaZderMind 17 Oct 05, 2022
A curated list of programmatic weak supervision papers and resources

A curated list of programmatic weak supervision papers and resources

Jieyu Zhang 118 Jan 02, 2023
Contains code for the paper "Vision Transformers are Robust Learners".

Vision Transformers are Robust Learners This repository contains the code for the paper Vision Transformers are Robust Learners by Sayak Paul* and Pin

Sayak Paul 103 Jan 05, 2023
The mini-AlphaStar (mini-AS, or mAS) - mini-scale version (non-official) of the AlphaStar (AS)

A mini-scale reproduction code of the AlphaStar program. Note: the original AlphaStar is the AI proposed by DeepMind to play StarCraft II.

Ruo-Ze Liu 216 Jan 04, 2023
[NeurIPS'21] "AugMax: Adversarial Composition of Random Augmentations for Robust Training" by Haotao Wang, Chaowei Xiao, Jean Kossaifi, Zhiding Yu, Animashree Anandkumar, and Zhangyang Wang.

[NeurIPS'21] "AugMax: Adversarial Composition of Random Augmentations for Robust Training" by Haotao Wang, Chaowei Xiao, Jean Kossaifi, Zhiding Yu, Animashree Anandkumar, and Zhangyang Wang.

VITA 112 Nov 07, 2022