Advanced Genetic Programming Techniques for Machine Learning - A Comparative Analysis with State-of-the-Art Automated Machine Learning Methods in Imbalanced Binary Classification

Aug 2, 2026

Upload by Logic

The thesis examines the rising use of advanced genetic programming concepts and automated machine learning tools while addressing the difficulty for applicants to evaluate emerging methods against established approaches. It compares two genetic programming techniques—differentiable Cartesian genetic programming for artificial neural networks and geometric semantic genetic programming—against Auto-Keras, Auto-PyTorch, and Auto-sklearn. Experiments run on 20 benchmark datasets focusing on average and maximum performance for imbalanced binary classification, then the best method is applied to real-world fraud detection. The goal is to determine whether the novel techniques can compete and identify the best-performing overall approach.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages72

Reading Time3 h 1 min

AccessFree to download

Introduction

Literature review

Automated machine learning

Genetic programming

Artificial neural networks

Methodology

Application to benchmark datasets

Application to fraud detection

Evaluation metrics

Setups of the techniques under analysis

Results and discussion

Master Degree Program in Data Science and Advanced Analytics MDSAA ADVANCED GENETIC PROGRAMMING TECHNIQUES FOR MACHINE LEARNING A comparative analysis with state-of-the-art automated machine learning methods in the context of imbalanced binary classification Franz Michael Frank Dissertation presented as partial requirement for obtaining the Master Degree in Data Science and Advanced Analytics NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa ADVANCED GENETIC PROGRAMMING TECHNIQUES FOR MACHINE LEARNING A comparative analysis with state-of-the-art automated machine learning methods in the context of imbalanced binary classification by Franz Michael Frank Dissertation presented as partial requirement for obtaining the Master degree in Data Science Advanced Analytics, with a Specialization in Data Science Supervisor: Prof. Fernando JosĠ Ferreira Lucas BaĕĆo 10 2022 STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Rules of Conduct and Code of Honor from the NOVA Information Management School. Lisbon, 13th of October 2022 ABSTRACT The number of available machine learning methods and tools is increasing rapidly, with one recent trend being the usage of advanced genetic programming concepts and automated machine learning tools. However, through the rising number of upcoming innovations, it has become a challenge for machine learning applicants to keep up with all the new opportunities and to identify their potentials. While emerging methods are typically compared to conventional standard machine learning algorithms upon their initial introduction, research is still scarce on comparisons of the performances between the new concepts themselves. Therefore, this thesis provides a comparative analysis of two novel genetic programming techniques, differentiable Cartesian genetic programming for artificial neural networks and geometric semantic genetic programming, alongside three state-of-the-art automated machine learning tools, Auto-Keras, Auto-PyTorch and Auto-sklearn, with regard to their relative performances in the machine learning subfield of imbalanced binary classification. In this analysis, the five methods are tested against each other on 20 benchmark datasets, primarily regarding their average and maximum performance, and subsequently the most successful technique is applied to the real-world problem of fraud detection. The purpose of this thesis is not only to familiarize machine learning users with these methods, but above all to determine whether the novel genetic programming techniques can compete with the more established automated machine learning tools, and to identify the overall best performing method. KEYWORDS Genetic Programming; Automated Machine Learning; Imbalanced Binary Classification INDEX 1. Introduction .................................................................................................................. 1 2. Literature review .......................................................................................................... 3 2.1. Automated machine learning ................................................................................3 2.1.1. Auto-Keras ......................................................................................................4 2.1.2. Auto-PyTorch ..................................................................................................4 2.1.3. Auto-sklearn ...................................................................................................4

FAQs

Which methods are compared in the thesis?

The thesis compares two genetic programming techniques—differentiable Cartesian genetic programming for artificial neural networks and geometric semantic genetic programming—alongside Auto-Keras, Auto-PyTorch, and Auto-sklearn.

How are the methods evaluated?

All five methods are tested against each other on 20 benchmark datasets, primarily using average and maximum performance metrics for imbalanced binary classification.

How is the best-performing method used after benchmarking?

The most successful technique from the benchmark comparison is applied to the real-world problem of fraud detection.

Aug 2, 2026

Upload by Logic