Genetic Programming in the Context of Natural Computing

Table of Contents

1 Natural Computing

1.1 Natural Hardware

1.1.1 Quantum Computing

1.1.2 DNA Computing

1.1.3 Optical Computing

1.2 Simulation of Nature

1.2.1 Fractal Geometry

1.2.2 Chaos Theory

1.2.3 Artificial Life

1.2.4 Other methods

1.3 Natural Methods = Bio-Inspired Computing

2 Bio-Inspired Computing

2.1 Evolutionary Computation

2.2 Neural Networks

2.2.1 The Blue Brain Project

2.3 Artificial immune systems

2.4 Swarm Intelligence

2.5 Other Types of Bio-inspired Computing

3 Systematic Literature Analysis

3.1 Automatic Time-Series-Search with a Perl-Script

3.2 Treatment of raw data

3.3 Conclusion

4 Genetic Algorithms

4.1 Evolutionary Strategies (ES)

4.1.1 The Process of Optimization

4.2 Genetic Algorithms (GA)

4.2.1 GA compared to Real Life

4.3 Integration of ES in GA

5 Genetic Programming

5.1 When is it suggested to use GP?

5.2 Basic Algorithm of GP

5.3 Initial Population

5.3.1 Terminal and Function Set

5.3.2 Some Constrains

5.3.3 Growing Trees

5.3.4 Seeding

5.4 Execute Programs and Ascertain their Fitness

5.4.1 Measuring the Quality

5.4.2 Multi-Objective Problems

5.4.3 Achieving Fitness

5.5 Creation of next Generation

5.5.1 Reproduction

5.5.2 Mutation

5.5.3 Crossover

5.5.4 Architecture Altering

5.5.5 Automatically Defined Functions (ADF)

5.6 End of Execution

5.7 Hall of Fame

6 Summary and Conclusion

Objectives and Research Themes

This thesis examines the field of Bio-Inspired Computing, specifically focusing on Genetic Programming (GP) as a heuristic optimization method. The central research question investigates how natural computing phenomena can be effectively applied to computational problem-solving, exploring the criteria for selecting genetic programming and the methodologies used to optimize computer programs through iterative loops.

• Systematic literature analysis of the last 30 years in Natural Computing.
• Comprehensive overview of Bio-Inspired Computing disciplines (Neural Networks, Swarm Intelligence, Immune Systems).
• Technical implementation and optimization of Genetic Algorithms and Genetic Programming.
• Practical methodologies for program evolution, including fitness determination and structural alteration.
• Assessment of problem-solving efficacy through real-world examples and the "Hall of Fame".

Excerpt from the Book

Summary of Chapters

1 Natural Computing: Provides an introductory overview of the field, covering Natural Hardware, nature simulation, and the categorization of bio-inspired methods.

2 Bio-Inspired Computing: Details specific nature-inspired disciplines including evolutionary computation, neural networks, artificial immune systems, and swarm intelligence.

3 Systematic Literature Analysis: Describes the methodology used to analyze 30 years of scientific literature, including script-based data collection and quantitative analysis techniques.

4 Genetic Algorithms: Explains the foundations of Genetic Algorithms, Evolutionary Strategies, and the adaptation of natural biological principles into computational optimization.

5 Genetic Programming: Focuses on the core of the thesis, covering the basic algorithm of GP, population initialization, fitness evaluation, and operator-based program evolution.

6 Summary and Conclusion: Synthesizes the findings of the research, evaluating the practical impact of the explored optimization techniques.

Keywords

Genetic Programming, Program Optimization, Bio-Inspired Computation, Literature Analysis, Fitness, Crossover, Mutation, Selection, Neural Networks, Evolutionary Strategies, Swarm Intelligence, Natural Computing, Artificial Immune Systems, Quantum Computing, Data Analysis

Frequently Asked Questions

What is the core focus of this thesis?

The work provides a comprehensive examination of Natural Computing, with a specific focus on the heuristic optimization of computer programs through Bio-Inspired methods, particularly Genetic Programming.

Which scientific fields does the author cover?

The thesis explores Natural Hardware (Quantum/DNA computing), nature simulation (Fractal geometry), and various Bio-Inspired Computing branches like neural networks and evolutionary computation.

What is the primary research goal?

The main goal is to evaluate the applicability of genetic programming in solving complex computational problems and to provide a practical guideline for its implementation and optimization.

Which methodologies are employed for the research?

The author performs a systematic quantitative literature analysis using custom Perl scripts to collect and process data from search engines, followed by a qualitative examination of GA/GP techniques.

What topics are discussed in the main body regarding Genetic Programming?

The main body details the basic GP algorithm, initial population generation, the importance of terminal and function sets, and the various operators like mutation and crossover used to evolve program trees.

Which keywords characterize this research best?

The work is best defined by terms such as Genetic Programming, Program Optimization, Bio-Inspired Computation, Crossover, Mutation, and Evolutionary Strategies.

How are "Bloat" and "Intron Elimination" handled in GP?

The author explains "Bloat" as a common issue in GP and introduces structural alteration techniques like Intron Elimination and Code Simplification to maintain program efficiency.

What is the "Blue Brain Project" and why is it included?

The Blue Brain Project is used as a prominent example of biologically accurate artificial neural network simulation, highlighting the ambition of modern computational neuroscience research.

How does the author define the quality of a program?

The author distinguishes between quality (a metric of how well a program solves a specific problem) and fitness (the probability of a program being selected for the next evolutionary generation).

What role does the "Hall of Fame" play in the thesis?

It highlights real-world, award-winning applications where GP-evolved programs have successfully outperformed conventionally developed human solutions in specific computational tasks.