Algorithm evolution with internal reinforcement for signal understanding

Automated program evolution has existed in some form for almost forty years. Signal understanding (e.g., signal classification) has been a scientific concern for longer than that. Generating a general machine learning signal understanding system has more recently attracted considerable research interest. First, this thesis defines and creates a general machine learning approach for signal understanding independent of the signal's type and size. This is accomplished through an evolutionary strategy of signal understanding programs that is an extension of genetic programming. Second, this thesis introduces a suite of sub-mechanisms that increase the power of genetic programming and contribute to the understanding of the learning technique developed. The central algorithmic innovation of this thesis is the process by which a novel principled credit-blame assignment is introduced and incorporated into the evolution of algorithms, thus improving the evolutionary process. This principled credit-blame assignment is done through a new program representation called neural programming and applied through a set of principled processes collectively called internal reinforcement in neural programming . This thesis concentrates on these algorithmic innovations in real world signal domains where the signals are typically large and/or poorly understood. This evolutionary learning of algorithms takes place in PADO , a system developed in this thesis for “parallel algorithm discovery and orchestration” and as a demonstrably effective strategy for divide-and-conquer in signal classification domains. This thesis includes an extensive empirical evaluation of the techniques developed in a rich variety of real-world signals. The results obtained demonstrate, among other things, the effectiveness of principled credit-blame assignment in algorithm evolution. This work is unique in three aspects. No other currently existing system can learn to classify or otherwise “symbolize” signals with no space or size penalties for the signal's size or type. No other system based on genetic programming currently exists that purposefully generates and orchestrates a variety of experts along problem specific lines. And, most centrally, the thesis introduces the first analytically sound mechanism for explaining and reinforcing specific parts of an evolving program. The goal of this thesis is to argue, explain, and demonstrate how representation and search are intimately connected in evolutionary computation and to address these dual concerns in the context of the evolution of Turing complete programs. Ideally, this thesis will inspire future research in this same area and along similar lines.

[1]  Shumeet Baluja,et al.  Genetic Algorithms and Explicit Search Statistics , 1996, NIPS.

[2]  Rodney A. Brooks,et al.  Elephants don't play chess , 1990, Robotics Auton. Syst..

[3]  David H. Wolpert,et al.  Stacked generalization , 1992, Neural Networks.

[4]  Frédéric Gruau,et al.  Cellular Encoding for interactive evolutionary robotics , 1996 .

[5]  D. E. Rumelhart,et al.  Learning internal representations by back-propagating errors , 1986 .

[6]  Astro Teller,et al.  Turing completeness in the language of genetic programming with indexed memory , 1994, Proceedings of the First IEEE Conference on Evolutionary Computation. IEEE World Congress on Computational Intelligence.

[7]  Sandip Sen,et al.  Strongly Typed Genetic Programming in Evolving Cooperation Strategies , 1995, ICGA.

[8]  K. Aikawa Speech recognition using time-warping neural networks , 1991, Neural Networks for Signal Processing Proceedings of the 1991 IEEE Workshop.

[9]  Rodney A. Brooks,et al.  From earwigs to humans , 1997, Robotics Auton. Syst..

[10]  Peter J. Angeline Evolutionary algorithms and emergent intelligence , 1993 .

[11]  M. Veloso Program Evolution for Data Mining , 1995 .

[12]  William B. Langdon Data structures and genetic programming , 1995 .

[13]  R. Collins Studies in artificial evolution , 1992 .

[14]  Jérôme Jaffré,et al.  A Symbolic Code Generator for Parameter Estimation , 1996 .

[15]  Kenneth E. Kinnear,et al.  Generality and Difficulty in Genetic Programming: Evolving a Sort , 1993, ICGA.

[16]  Astro Teller,et al.  Algorithm evolution for face recognition: what makes a picture difficult , 1995, Proceedings of 1995 IEEE International Conference on Evolutionary Computation.

[17]  Thomas S. Huang,et al.  Evolvable 3D modeling for model-based object recognition systems , 1994 .

[18]  John R. Koza,et al.  Classifying proteins as extracellular using programmatic motifs and genetic programming , 1998, 1998 IEEE International Conference on Evolutionary Computation Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98TH8360).

[19]  Thomas Beck,et al.  The if-problem in automatic differentiation , 1994 .

[20]  John R. Koza,et al.  Genetic programming 2 - automatic discovery of reusable programs , 1994, Complex adaptive systems.

[21]  Scott Brave,et al.  The evolution of memory and mental models using genetic programming , 1996 .

[22]  Robert E. Smith,et al.  Classifier system renaissance: new analogies, new directions , 1996 .

[23]  William B. Langdon Evolving Data Structures with Genetic Programming , 1995, ICGA.

[24]  Peter J. Angeline,et al.  Algorithm Discovery using the Genetic Programming Paradigm: Extracting Low-Contrast Curvilinear Features from SAR Images of Arctic Ice , 1996 .

[25]  Peter Nordin,et al.  Genetic programming - An Introduction: On the Automatic Evolution of Computer Programs and Its Applications , 1998 .

[26]  Nichael Lynn Cramer,et al.  A Representation for the Adaptive Generation of Simple Sequential Programs , 1985, ICGA.

[27]  Dean A. Pomerleau,et al.  Neural Network Perception for Mobile Robot Guidance , 1993 .

[28]  Astro Teller,et al.  PADO: a new learning architecture for object recognition , 1997 .

[29]  Astro Teller,et al.  Evolving programmers: the co-evolution of intelligent recombination operators , 1996 .

[30]  Frédéric Gruau,et al.  Genetic micro programming of neural networks , 1994 .

[31]  N. Littlestone Learning Quickly When Irrelevant Attributes Abound: A New Linear-Threshold Algorithm , 1987, 28th Annual Symposium on Foundations of Computer Science (sfcs 1987).

[32]  David Andre,et al.  Automatically defined features: the simultaneous evolution of 2-dimensional feature detectors and an algorithm for using them , 1994 .

[33]  Peter Nordin,et al.  Speech Sound Discrimination with Genetic Programming , 1998, EuroGP.

[34]  Scott Brave,et al.  Evolving deterministic finite automata using cellular encoding , 1996 .

[35]  R. Poli Genetic programming for image analysis , 1996 .

[36]  Peter J. Angeline,et al.  Two self-adaptive crossover operators for genetic programming , 1996 .

[37]  David B. Fogel,et al.  Alternative Neural Network Training Methods , 1995, IEEE Expert.

[38]  Peter J. Angeline,et al.  Evolutionary Module Acquisition , 1993 .

[39]  Claude Sammut,et al.  Automatic Speaker Recognition: An Application of Machine Learning , 1995, ICML.

[40]  Stewart W. Wilson Hierarchical Credit Allocation in a Classifier System , 1987, IJCAI.

[41]  Jeffrey D. Ullman,et al.  Introduction to Automata Theory, Languages and Computation , 1979 .

[42]  Lee Spector,et al.  Evolving teamwork and coordination with genetic programming , 1996 .

[43]  Vidroha Debroy,et al.  Genetic Programming , 1998, Lecture Notes in Computer Science.

[44]  P. Aloy,et al.  Relation between amino acid composition and cellular location of proteins. , 1997, Journal of molecular biology.

[45]  Avrim Blum,et al.  Empirical Support for Winnow and Weighted-Majority Based Algorithms: Results on a Calendar Scheduling Domain , 1995, ICML.

[46]  Lawrence J. Fogel,et al.  Artificial Intelligence through Simulated Evolution , 1966 .

[47]  Justinian P. Rosca,et al.  Discovery of subroutines in genetic programming , 1996 .

[48]  P. Ross,et al.  An adverse interaction between crossover and restricted tree depth in genetic programming , 1996 .

[49]  Rich Caruana,et al.  Removing the Genetics from the Standard Genetic Algorithm , 1995, ICML.

[50]  Astro Teller,et al.  Evolving Team Darwin United , 1998, RoboCup.

[51]  Fachbereich Informatik,et al.  Programmatic Compression of Images and Sound , 1996 .

[52]  Paul A. Viola Feature-Based Recognition of Objects , 1993 .

[53]  John H. Holland,et al.  Cognitive systems based on adaptive algorithms , 1977, SGAR.

[54]  A. Bairoch,et al.  The SWISS-PROT protein sequence data bank: current status. , 1994, Nucleic acids research.

[55]  Walter Alden Tackett,et al.  Recombination, selection, and the genetic construction of computer programs , 1994 .

[56]  Takeo Kanade,et al.  Human Face Detection in Visual Scenes , 1995, NIPS.

[57]  Geoffrey E. Hinton,et al.  Phoneme recognition using time-delay neural networks , 1989, IEEE Trans. Acoust. Speech Signal Process..

[58]  Peter L. Bartlett,et al.  The Canonical Distortion Measure in Feature Space and 1-NN Classification , 1997, NIPS.

[59]  Astro Teller,et al.  A study in program response and the negative effects of introns in genetic programming , 1996 .

[60]  Sebastian Thrun,et al.  Learning One More Thing , 1994, IJCAI.

[61]  L. Altenberg The evolution of evolvability in genetic programming , 1994 .

[62]  Roland Olsson,et al.  Inductive Functional Programming Using Incremental Program Transformation , 1995, Artif. Intell..

[63]  Tomaso A. Poggio,et al.  Example-Based Learning for View-Based Human Face Detection , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[64]  Astro Teller,et al.  The evolution of mental models , 1994 .

[65]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[66]  F. H. Bennett,et al.  Discovery by genetic programming of a cellular automata rule that is better than any known rule for the majority classification problem , 1996 .

[67]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[68]  Rodney A. Brooks,et al.  A Robust Layered Control Syste For A Mobile Robot , 2022 .

[69]  Chris Watkins,et al.  Learning from delayed rewards , 1989 .

[70]  John R. Koza,et al.  Automated WYWIWYG design of both the topology and component values of electrical circuits using genetic programming , 1996 .

[71]  Translator EnvironmentChristian Bischof ADIFOR : Automatic Di erentiationin a Source , 2007 .

[72]  L. Darrell Whitley,et al.  Cellular Encoding Applied to Neurocontrol , 1995, ICGA.

[73]  David Andre,et al.  The Evolution of Agents that Build Mental Models and Create Simple Plans Using Genetic Programming , 1995, ICGA.

[74]  R. Poli,et al.  Discovery of Symbolic, Neuro-Symbolic and Neural Networks with Parallel Distributed Genetic Programming , 1997, ICANNGA.

[75]  Kalyanmoy Deb,et al.  Messy Genetic Algorithms: Motivation, Analysis, and First Results , 1989, Complex Syst..

[76]  Walter Alden Tackett,et al.  Genetic Programming for Feature Discovery and Image Discrimination , 1993, ICGA.

[77]  J. K. Kinnear,et al.  Advances in Genetic Programming , 1994 .

[78]  Astro Teller,et al.  PADO: Learning Tree Structured Algorithms for Orchestration into an Object Recognition System , 1995 .

[79]  Ken Sharman,et al.  Evolving signal processing algorithms by genetic programming , 1995 .

[80]  Shumeet Baluja Artificial Neural Network Evolution: Learning to Steer a Land Vehicle , 1995 .

[81]  O. G. Selfridge,et al.  Pandemonium: a paradigm for learning , 1988 .

[82]  David B. Fogel,et al.  An Evolutionary Programming Approach to Self-Adaptation on Finite State Machines , 1995, Evolutionary Programming.

[83]  Peter Nordin,et al.  Evolutionary program induction of binary machine code and its applications , 1997 .

[84]  M. Turk,et al.  Eigenfaces for Recognition , 1991, Journal of Cognitive Neuroscience.

[85]  A. Waibel,et al.  Multi-speaker/speaker-independent architectures for the multi-state time delay neural network , 1993, 1993 IEEE International Conference on Acoustics, Speech, and Signal Processing.