Winston Ewert

Evolutionary synthesis of nand logic: Dissecting a digital organism

According to conservation of information theorems, performance of an arbitrarily chosen search, on average, does no better than blind search. Domain expertise and prior knowledge about search space structure or target location is therefore essential in crafting the search algorithm. The effectiveness of a given algorithm can be measured by the active information introduced to the search. We illustrate this by identifying sources of active information in Avida, a software program designed to search for logic functions using nand gates. Avida uses stair step active information by rewarding logic functions using a smaller number of nands to construct functions requiring more. Removing stair steps deteriorates Avidas performance while removing deleterious instructions improves it. Some search algorithms use prior knowledge better than others. For the Avida digital organism, a simple evolutionary strategy generates the Avida target in far fewer instructions using only the prior knowledge available to Avida.

Evolutionary Inversion of Swarm Emergence Using Disjunctive Combs Control

Given simple agent rules, a swarms emergent behavior can be difficult to predict. The inverse problem is even more difficult: Given a desired emergent behavior, what are the rules by which swarm agents should operate? Disjunctive fuzzy control is proposed as a method to model swarm agents. Compared to more commonly used conjunctive fuzzy control such as that proposed by Mamdani, disjunctive fuzzy control is robustly fault tolerant and disjointly connected. Swarms are inherently disjunctive. Instead of agents working in coordination with one another, each swarm agent contributes individually to the result. The disjunctive attribute can also be applied at the sensor level for each individual agent. Disjunctive control allows adaptation of the describing membership function, as is commonly done in conjunctive control. The inversion process is illustrated with numerous simulation examples, including a predator-prey game, gang warfare, and escaping agents. The swarm is instructed what to do but not how to do it. Imposition of fitness constraints and repeated generations of evolutionary molding of agent performance can then result in unexpected emergent behaviors of the swarm, e.g., use of decoys, self-sacrifice, flanking maneuvers, and shielding of the weak.

Efficient per query information extraction from a Hamming oracle

Computer search often uses an oracle to determine the value of a proposed problem solution. Information is extracted from the oracle using repeated queries. Crafting a search algorithm to most efficiently extract this information is the job of the programmer. In many instances this is done using the programmers experience and knowledge of the problem being solved. For the Hamming oracle, we have the ability to assess the performance of various search algorithms using the currency of query count. Of the search procedures considered, blind search performs the worst. We show that evolutionary algorithms, although better than blind search, are a relatively inefficient method of information extraction. An algorithm methodically establishing and tracking the frequency of occurrence of alphabet characters performs even better. We also show that a search for the search for an optimal tree search, as suggested by our previous work, becomes computationally intensive.

A Vivisection of the ev Computer Organism: Identifying Sources of Active Information

ev is an evolutionary search algorithm proposed to simulate biological evolution. As such, researchers have claimed that it demonstrates that a blind, unguided search is able to generate new information. However, analysis shows that any non-trivial computer search needs to exploit one or more sources of knowledge to make the search successful. Search algorithms mine active information from these resources, with some search algorithms performing better than others. We illustrate these principles in the analysis of ev . The sources of knowledge in ev include a Hamming oracle and a perceptron structure that predisposes the search towards its target. The original ev uses these resources in an evolutionary algorithm. Although the evolutionary algorithm finds the target, we demonstrate a simple stochastic hill climbing algorithm uses the resources more efficiently.

Unexpected emergent behaviors from elementary swarms

Swarms are collections of loosely coupled distinct agents each following simple rules. Swarms do not use central coordination AND individual agents need not be aware of the swarms overall function. As each agent performs its task, the swarm collective can display unusual and unexpected emergent behaviors. For those swarms defying analytic evaluation, simulation remains as the only method to reveal emergent behavior. A number of swarms are simulated each with no more than simple rules to follow. Each simulation reveals an interesting and often surprising emergent behavior. Termites clear areas and stack wood chips, gnats naturally confine themselves to swarm inside a circle of fixed area and sand piles develop instabilities and avalanche. Possibly the most interesting simulation is predator swarms pursuing swarms of prey in a game we dub bullies and dweebs. Individual bullies can be ineffective in killing dweebs, for example, whereas a mob of bullies can be highly effective. Addition of stochastic component to dweeb motion in a swarm is essential for prolonging dweeb life. The swarms are illustrated using screen shots of the swarm dynamics. More interesting and insightful videos of the swarming are available on NeoSwarm.com.

Climbing the Steiner Tree--Sources of Active Information in a Genetic Algorithm for Solving the Euclidean Steiner Tree Problem

Genetic algorithms are widely cited as demonstrating the power of natural selection to produce biological complexity. In particular, the success of such search algorithms is said to show that intelligent design has no scientific value. Despite their merits, genetic algorithms establish nothing of the sort. Such algorithms succeed not through any intrinsic prop- erty of the search algorithm, but rather through incorporating sources of information derived from the programmer’s prior knowledge. A genetic algorithm used to defend the efficacy of natural selection is Thomas’s Steiner tree algorithm. This paper tracks the various sources of information incorporated into Thomas’s algorithm. Rather than creating informa- tion from scratch, the algorithm incorporates resident information by restricting the set of solutions considered, introducing selection skew to increase the power of selection, and adopting a structure that facilitates fortuitous crossover. Thomas’s algorithm, far from exhibiting the power of natural selection, merely demonstrates that an intelligent agent, in this case a human programmer, possesses the ability to incorporate into such algorithms the information necessary for successful search.

On the improbability of algorithmic specified complexity

An event with low probability is unlikely to happen, but events with low probability happen all of the time. This is because many distinct low probability events can have a large combined probability. However, some low probability events can be seen to follow an independent pattern. Algorithmic specified complexity (ASC) measures the degree to which an event is improbable and follows a pattern. We show a bound on the probability of obtaining a particular value of algorithmic specified complexity. Consequently we can say that high ASC objects are improbable.

A General Theory of Information Cost Incurred by Successful Search

This paper provides a general framework for understanding targeted search. It begins by defining the search matrix, which makes explicit the sources of information that can affect search progress. The search matrix enables a search to be represented as a probability measure on the original search space. This representation facilitates tracking the information cost incurred by successful search (success being defined as finding the target). To categorize such costs, various information and efficiency measures are defined, notably, active information. Conservation of information characterizes these costs and is precisely formulated via two theorems, one restricted (proved in previous work of ours), the other general (proved for the first time here). The restricted version assumes a uniform probability search baseline, the general, an arbitrary probability search baseline. When a search with probability q of success displaces a baseline search with probability p of success where q > p, conservation of information states that raising the probability of successful search by a factor of q/p(>1) incurs an information cost of at least log(q/p). Conservation of information shows that information, like money, obeys strict accounting principles.

Tierra: The Character of Adaptation

Tierra is a digital simulation of evolution for which the stated goal was the development of openended complexity and a digital “ Cambrian Explosion.” However, Tierra failed to produce such a result. A closer inspection of Tierran evolution’s adaptations show very few instances of adaptation through the production of new information. Instead, most changes result from removing or rearranging the existing pieces within a Tierra program. The open-ended development of complexity depends on the ability to generate new information, but this is precisely what Tierra struggles to do. The character of Tierran adaptation does not allow for open-ended complexity but is similar to the character of adaptations found in the biological world.

The Dependency Graph of Life

The hierarchical classification of life has been claimed as compelling evidence for universal common ancestry. However, research has uncovered much data which is not congruent with the hierarchical pattern. Nevertheless, biological data resembles a nested hierarchy sufficiently well to require an explanation. While many defenders of intelligent design dispute common descent, no alternative account of the approximate nested hierarchy pattern has been widely adopted. We present the dependency graph hypothesis as an alternative explanation, based on the technique used by software developers to reuse code among different software projects. This hypothesis postulates that different biological species share modules related by a dependency graph. We evaluate several predictions made by this model about both biological and synthetic data, finding them to be fulfilled.