why-functional-program-synthesis-matters

所属分类:人工智能/神经网络/深度学习
开发工具:Haskell
文件大小:6616KB
下载次数:0
上传日期:2022-07-06 15:19:15
上 传 者sh-1993
说明:  在遗传编程领域
(in the realm of genetic programming)

文件列表:
CC-BY.markdown (17067, 2022-06-24)
PonyGE2-LICENSE (35149, 2022-06-24)
compile_results.py (2738, 2022-06-24)
datasets (0, 2022-06-24)
datasets\gpsbs (0, 2022-06-24)
datasets\gpsbs\hs (0, 2022-06-24)
datasets\gpsbs\hs\10_Wallis_Pi (0, 2022-06-24)
datasets\gpsbs\hs\10_Wallis_Pi\10_Wallis_Pi-Example-Test.hs (2338, 2022-06-24)
datasets\gpsbs\hs\10_Wallis_Pi\10_Wallis_Pi-Example-Train.hs (2338, 2022-06-24)
datasets\gpsbs\hs\10_Wallis_Pi\Test.txt (583, 2022-06-24)
datasets\gpsbs\hs\10_Wallis_Pi\Train.txt (1712, 2022-06-24)
datasets\gpsbs\hs\11_String_Length_Backwards (0, 2022-06-24)
datasets\gpsbs\hs\11_String_Length_Backwards\11_String_Length_Backwards-Example-Test.hs (808608, 2022-06-24)
datasets\gpsbs\hs\11_String_Length_Backwards\11_String_Length_Backwards-Example-Train.hs (77781, 2022-06-24)
datasets\gpsbs\hs\11_String_Length_Backwards\Test.txt (807019, 2022-06-24)
datasets\gpsbs\hs\11_String_Length_Backwards\Train.txt (76132, 2022-06-24)
datasets\gpsbs\hs\12_Last_Index_Of_Zero (0, 2022-06-24)
datasets\gpsbs\hs\12_Last_Index_Of_Zero\12_Last_Index_Of_Zero-Example-Test.hs (109792, 2022-06-24)
datasets\gpsbs\hs\12_Last_Index_Of_Zero\12_Last_Index_Of_Zero-Example-Train.hs (13176, 2022-06-24)
datasets\gpsbs\hs\12_Last_Index_Of_Zero\Test.txt (108496, 2022-06-24)
datasets\gpsbs\hs\12_Last_Index_Of_Zero\Train.txt (11880, 2022-06-24)
datasets\gpsbs\hs\13_Vector_Average (0, 2022-06-24)
datasets\gpsbs\hs\13_Vector_Average\13_Vector_Average-Example-Test.hs (508221, 2022-06-24)
datasets\gpsbs\hs\13_Vector_Average\13_Vector_Average-Example-Train.hs (50838, 2022-06-24)
datasets\gpsbs\hs\13_Vector_Average\Test.txt (506751, 2022-06-24)
datasets\gpsbs\hs\13_Vector_Average\Train.txt (49367, 2022-06-24)
datasets\gpsbs\hs\14_Count_Odds (0, 2022-06-24)
datasets\gpsbs\hs\14_Count_Odds\14_Count_Odds-Example-Test.hs (267698, 2022-06-24)
datasets\gpsbs\hs\14_Count_Odds\14_Count_Odds-Example-Train.hs (22747, 2022-06-24)
datasets\gpsbs\hs\14_Count_Odds\Test.txt (266501, 2022-06-24)
datasets\gpsbs\hs\14_Count_Odds\Train.txt (21550, 2022-06-24)
datasets\gpsbs\hs\15_Mirror_Image (0, 2022-06-24)
datasets\gpsbs\hs\15_Mirror_Image\15_Mirror_Image-Example-Test.hs (279901, 2022-06-24)
datasets\gpsbs\hs\15_Mirror_Image\15_Mirror_Image-Example-Train.hs (21237, 2022-06-24)
datasets\gpsbs\hs\15_Mirror_Image\Test.txt (278662, 2022-06-24)
datasets\gpsbs\hs\15_Mirror_Image\Train.txt (19998, 2022-06-24)
datasets\gpsbs\hs\16_Super_Anagrams (0, 2022-06-24)
... ...

# Introduction This repository contains the all relevant materials to perform the experiments presented in [Why Functional Program Synthesis Matters (In The Realm of Genetic Programming)](https://dl.acm.org/doi/10.1145/3520304.3534045) [1]. The experiments are carried out using the [PonyGE2](https://github.com/PonyGE/PonyGE2) GE implementation [2] in the CFG-GP form. The problems are from the [General Program Synthesis Benchmark Suite](https://thelmuth.github.io/GECCO_2015_Benchmarks_Materials/) [3] (note, the data was taken from the PonyGE2 repository). A persistent copy of the project as used to generate the results in [1] can be found at the Zenodo record [here](https://zenodo.org/record/***99027). # Requirements We assume a unix OS, the code was run/tested on Pop_OS!-20.04 (Ubuntu derivative). Python 3.5 (or higher) is required, Python 3.8.10 was used for the reported results. The Python packages required are: matplotlib, numpy, scipy, scikit-learn (sklearn), pandas, which can be installed as follows via your favourite terminal emulator: $ pip install -r requirements.txt GHC (Glorious Glasgow Haskell Compilation System) version 8.10.7 was used for the reported results. GHC can be installed using the following command and following the instructions: $ curl --proto '=https' --tlsv1.2 -sSf https://get-ghcup.haskell.org | sh # Running the experiments performed You can run all the experiments by executing the experiment script in the src folder $ cd src $ sh experiments.sh This will run all the experiments presented in the paper. The results will be written to the automatically generated results folder. Possible problems: make sure 'runhaskell' is available on your system path. # License This work is published under a Creative Commons 4.0 CC-BY [license](CC-BY.markdown). The markdown version of the license was taken from this great [project](https://github.com/idleberg/Creative-Commons-Markdown). For PonyGE2 see [PonyGE2-LICENSE](PonyGE2-LICENSE) or the [PonyGE2](https://github.com/PonyGE/PonyGE2) repo. # References 1. Garrow, F., Lones, M.A., and Stewart, R., Why Functional Program Synthesis Matters (In The Realm of Genetic Programming). GECCO '22: Proceedings of the Genetic and Evolutionary Computation Conference Companion. July 2022. ACM. 2. Fenton, M., McDermott, J., Fagan, D., Forstenlechner, S., Hemberg, E., and O'Neill, M. PonyGE2: Grammatical Evolution in Python. arXiv preprint, arXiv:1703.08535, 2017. 3. T. Helmuth and L. Spector. General Program Synthesis Benchmark Suite. In GECCO '15: Proceedings of the Genetic and Evolutionary Computation Conference. July 2015. ACM.

近期下载者

相关文件


收藏者