No abstract

Evolutionary Algorithms (EA) approach the genotype-phenotype relationship differently than does nature, and this discrepancy is a recurrent issue among researchers. Moreover, in spite of some performance improvements, it is a fact that biological knowledge has advanced faster than our ability to incorporate novel biological ideas into EAs. Recently, some researchers have started exploring computationally new comprehension of the multitude of the regulatory mechanisms that are fundamental in both processes of inheritance and of development in natural systems, by trying to include those mechanisms in the EAs. One of the first successful proposals was the Artificial Gene Regulatory Network (ARN) model, by Wolfgang Banzhaf. Soon after some variants of the ARN were tested. In this paper, we describe one of those, the Regulatory Network Computational Device, demonstrating experimentally its capabilities. The efficacy and efficiency of this alternative is tested experimentally using typical benchmark problems for Genetic Programming (GP) systems. We devise a modified factorial problem to investigate the use of feedback connections and the scalability of the approach. In order to gain a better understanding about the reasons for the improved quality of the results, we undertake a preliminary study about the role of neutral mutations during the evolutionary process.

Well logs are records of petro-physical data acquired along a borehole, providing direct information about what is in the subsurface. The data collected by logging wells can have significant economic consequences in oil and gas exploration, not only because it has a direct impact on the following decisions, but also due to the subsequent costs inherent to drilling wells, and the potential return of oil deposits. These logs frequently present gaps of varied sizes in the sensor recordings, that happen for diverse reasons. These gaps result in less information used by the interpreter to build the stratigraphic models, and consequently larger uncertainty regarding what will be encountered when the next well is drilled. The main goal of this work is to compare Gradient Tree Boosting, Random Forests, Artificial Neural Networks, and three algorithms of Linear Regression on the prediction of the gaps in well log data. Given the logs from a specific well, we use the intervals with complete information as the training data to learn a regression model of one of the sensors for that well. The algorithms are compared with each other using a few individual example wells with complete information, on which we build artificial gaps to cross validate the results. We show that the ensemble algorithms tend to perform significantly better, and that the results hold when addressing the different examples individually. Moreover, we performed a grid search over the ensembles parameters space, but did not find a statistically significant difference in any situation.

No abstract

The relationship between the genotype and the phenotype in Evolutionary Algorithms (EA) is a recurrent issue among researchers. Based on our current understanding of the multitude of the regulatory mechanisms that are fundamental in both processes of inheritance and of development in natural systems, some researchers start exploring computationally this new insight, including those mechanism in the EA. The Artificial Gene Regulatory (ARN) model, proposed by Wolfgang Banzhaf was one of the first tentatives. Following his seminal work some variants were proposed with increased capabilities. In this paper, we present another modification of this model, consisting in the use the regulatory network as a computational device where feedback edges are used. Using two classical benchmarks, the n-bit parity and the Fibonacci sequence problems, we show experimentally the effectiveness of the proposal.