Kosuke Imamura

Behavioral Diversity and a Probabilistically Optimal GP Ensemble

We propose N-version Genetic Programming (NVGP) as an ensemble method to enhance accuracy and reduce performance fluctuation of programs produced by genetic programming. Diversity is essential for forming successful ensembles. NVGP quantifies behavioral diversity of ensemble members and defines NVGP optimal as an ensemble that has independent fault occurrences among its members. We observed significant accuracy improvement by NVGP optimal ensembles when applied to a DNA segment classification problem.

The test vector problem and limitations to evolving digital circuits

How do we know the correctness of an evolved circuit? While Evolutionary Hardware is exhibiting its effectiveness, we argue that it is very difficult to design a large-scale digital circuit by conventional evolutionary techniques alone, if we are using a subset of the entire truth table for fitness evaluation. The test vector generation problem for testing VLSI (Very Large Scale Integration) suggests that there is no efficient way to determine a training set which assures full correctness of an evolved circuit.

N-Version Genetic Programming via Fault Masking

We introduce a new method, N-Version Genetic Programming (NVGP), for building fault tolerant software by building an ensemble of automatically generated modules in such a way as to maximize their collective fault masking ability. The ensemble itself is an example of n-version modular redundancy for fault tolerance, where the output of the ensemble is the most frequent output of n independent modules. By maximizing collective fault masking, NVGP approaches the fault tolerance expected from n version modular redundancy with independent faults in component modules. The ensemble comprises individual modules from a large pool generated with genetic programming, using operators that increase the diversity of the population. Our experimental test problem classified promoter regions in Escherichia coli DNA sequences. For this problem, NVGP reduced the number and variance of errors over single modules produced by GP, with statistical significance.

Potential application of training based computation to intrusion detection

Without detection of a network intrusion, a system is not capable of properly defending itself. Therefore, the first step in preserving system integrity is to detect whether or not the system is under attack. We initiated a research project that utilizes training based computation for network intrusion detection. The goal of this project is to defend the system from unknown attacks. Packet analysis approaches are effective at detecting known attacks, but fail at unknown attack detection. In order to protect the system from unknown attacks, we need to develop a classifier system which is independent of the signatures found in network packets. One of the promising ways to perform this classification is to profile kernel level activities. We apply a probabilistically optimal classifier ensemble method to monitor kernel activity, and ultimately to predict whether or not the system is under attack.

Square Wave Artificial Neuron (SWAN)

We developed a square wave based artificial neuron to take advantage of inexpensive and readily available Field Programmable Gate Array technologies. While conventional neurons require computationally intensive floating point arithmetic to determine the output, our artificial neuron converts inputs into square waves and the time that these waves require to produce a predetermined bit pattern is used to determine the output value. The behavior of this neuron was successfully simulated by software to demonstrate that the Square Wave Artificial Neuron is software implementable.