Genetic Evolutionary Programming Research Articles

Ensemble classifiers using multi-objective Genetic Programming for unbalanced data

Genetic Programming (GP) can be used to design effective classifiers due to its built-in feature selection and feature construction characteristics. Unbalanced data distributions affect the classification performance of GP classifiers. Some fitness functions have been proposed to solve the class imbalance problem of GP classifiers. However, with the evolution of GP, single-objective GP classifiers evaluated by a single fitness function have poor generalization ability. Moreover, using the best evolved GP classifier for decision-making can easily lead to the possibility of misclassification. In this paper, multi-objective GP is used to optimize multiple fitness functions including AUC approximation (Wmw), Distance (Dist), and Complexity to evolve ensemble classifiers, which jointly determines the class labels of unknown instances. Experiments on sixteen datasets show that our multi-objective GP can significantly improve classification performance compared with single-objective GP, and our proposed ensemble classifiers evolved by multi-objective GP can further improve the classification performance than the single best GP classifier. Comparisons with six GP-based and five traditional machine learning algorithms show that our proposed approaches can achieve significantly better classification performance on most cases.

Predictive Models for Forecasting Public Health Scenarios: Practical Experiences Applied during the First Wave of the COVID-19 Pandemic.

Background: Forecasting the behavior of epidemic outbreaks is vital in public health. This makes it possible to anticipate the planning and organization of the health system, as well as possible restrictive or preventive measures. During the COVID-19 pandemic, this need for prediction has been crucial. This paper attempts to characterize the alternative models that were applied in the first wave of this pandemic context, trying to shed light that could help to understand them for future practical applications. Methods: A systematic literature search was performed in standardized bibliographic repertoires, using keywords and Boolean operators to refine the findings, and selecting articles according to the main PRISMA 2020 statement recommendations. Results: After identifying models used throughout the first wave of this pandemic (between March and June 2020), we begin by examining standard data-driven epidemiological models, including studies applying models such as SIR (Susceptible-Infected-Recovered), SQUIDER, SEIR, time-dependent SIR, and other alternatives. For data-driven methods, we identify experiences using autoregressive integrated moving average (ARIMA), evolutionary genetic programming machine learning, short-term memory (LSTM), and global epidemic and mobility models. Conclusions: The COVID-19 pandemic has led to intensive and evolving use of alternative infectious disease prediction models. At this point it is not easy to decide which prediction method is the best in a generic way. Moreover, although models such as the LSTM emerge as remarkably versatile and useful, the practical applicability of the alternatives depends on the specific context of the underlying variable and on the information of the target to be prioritized. In addition, the robustness of the assessment is conditioned by heterogeneity in the quality of information sources and differences in the characteristics of disease control interventions. Further comprehensive comparison of the performance of models in comparable situations, assessing their predictive validity, is needed. This will help determine the most reliable and practical methods for application in future outbreaks and eventual pandemics.

Designing genetic programming classifiers with feature selection and feature construction

Due to the flexibility of Genetic Programming (GP), GP has been used for feature construction, feature selection and classifier construction. In this paper, GP classifiers with feature selection and feature construction are investigated to obtain simple and effective classification rules. During the construction of a GP classifier, irrelevant and redundant features affect the search ability of GP, and make GP easily fall into local optimum. This paper proposes two new GP classifier construction methods to restrict bad impact of irrelevant and redundant features on GP classifier. The first is to use a multiple-objective fitness function that decreases both classification error rate and the number of selected features, which is named as GPMO. The second is to first use a feature selection method, i.e., linear forward selection (LFS) to remove irrelevant and redundant features and then use GPMO to construct classifiers, which is named as FSGPMO. Experiments on twelve datasets show that GPMO and FSGPMO have advantages over GP classifiers with a single-objective fitness function named GPSO in term of classification performance, the number of selected features, time cost and function complexity. The proposed FSGPMO can achieve better classification performance than GPMO on higher dimension datasets, however, FSGPMO may remove potential effective features for GP classifier and achieve much lower classification performance than GPMO on some datasets. Compared with two other GP-based classifiers, GPMO can significantly improve the classification performance. Comparisons with other classification algorithms show that GPMO can achieve better or comparable classification performance on most selected datasets. Our proposed GPMO can achieve better performance than wrapper-based feature construction methods using GP on applications with insufficient instances. Further investigations show that bloat phenomena exists in the process of GP evolution and overfitting phenomena is not obvious. Moreover, the benefits of GP over other machine learning algorithms are discussed.

An ensemble genetic programming model for seasonal precipitation forecasting

Seasonal precipitation forecasting is one of the most challenging tasks in stochastic hydrology. This article proposes a new ensemble model, called EGP, to a season-ahead forecast of total seasonal precipitation. The EGP integrates evolutionary genetic programming (GP) and gene expression programming (GEP) techniques with multiple linear regression to increase forecasting accuracy of standalone GP and GEP models, while it secures their explicit structure. The EGP model was trained and validated using 88 years (1930–2017) of measured precipitation data from Muratpasa Station, Antalya, Turkey. The model performance was evaluated in terms of different statistical error measures and cross-validated with two other ensemble models as well as the state-of-the-art random forest developed in this study as the benchmark. The results showed that the proposed model can increase the forecasting accuracy of the best standalone GP and GEP models up to 30%. The EGP was also found to be superior to random forest, particularly in predicting low and high seasonal precipitation amount. This model is explicit, easy to evolve, and therefore, motivating to be used in practice.

Active control of vortex-induced vibration of a circular cylinder using machine learning

We demonstrate the use of high-fidelity computational fluid dynamics simulations in machine-learning based active flow control. More specifically, for the first time, we adopt the genetic programming (GP) to select explicit control laws, in a data-driven and unsupervised manner, for the suppression of vortex-induced vibration (VIV) of a circular cylinder in a low-Reynolds-number flow (Re = 100), using blowing/suction at fixed locations. A cost function that balances both VIV suppression and energy consumption for the control is carefully chosen according to the knowledge obtained from pure blowing/suction open-loop controls. By implementing reasonable constraints to VIV amplitude and actuation strength during the GP evolution, the GP-selected best ten control laws all point to suction-type actuation. The best control law suggests that the suction strength should be nonzero when the cylinder is at its equilibrium position and should increase nonlinearly with the cylinder’s transverse displacement. Applying this control law suppresses 94.2% of the VIV amplitude and achieves 21.4% better overall performance than the best open-loop controls. Furthermore, it is found that the GP-selected control law is robust, being effective in flows ranging from Re = 100 to 400. On the contrary, although the P-control can achieve similar performance as the GP-selected control at Re = 100, it deteriorates in higher Reynolds number flows. Although for demonstration purpose the chosen control problem is relatively simple, the training experience and insights obtained from this study can shed some light on future GP-based control of more complicated problems.

APPLICATION OF GENETIC PROGRAMMING TOOLS AS A MEANS OF SOLVING OPTIMIZATION PROBLEMS

Purpose. The article is devoted to the problem of practical application of genetic programming tools as a means of solving optimization problems and the use of genetic programming in various fields of activity. It is established that the evolution of genetic programming is directly related to the development of the genetic algorithm, it is also determined that with the passage of time a significant improvement in genetic programming has occurred. Since the advent of the genetic algorithm, many modifications and software implementations have appeared. This in turn led to the implementation of the genetic algorithm toolkit in software products, namely: specialized software, applications for mathematical and analytical packages, frameworks and libraries. The article reveals the significant impact of genetic programming in the areas of: quantum computing, electrical circuit design, etc. Not only advantages, but also disadvantages are considered, attention is also paid to methods of eliminating deficiencies by improving optimization methods and applying a genetic algorithm. Results. The analysis of the main directions of the practical use of genetic programming is carried out and tasks that can be effectively solved using this toolkit are outlined. Scientific novelty. It was determined that the improvement of optimization methods and the expansion of the use of genetic algorithms, stimulates the appearance of such software products on the market, simplifies the structure of software tools, designs the interface for working with a specific commercial user community, simplifies the command language, which allows the use of genetic programming tools circle of users with different levels of training.

Evaluation of Genetic Programs in Multiple Cases for Gait Classification and Recognition

The study of human gait is one of the interesting fields among several disciplines. This arena has attracted researchers and professionals not only from human physiology and neuro biology but has also fascinated scientists from socio-psychology and socio-informatics. The locomotion of humans is associated with lots of physical and biological aspect that might provide critical information related to stress, intentions, emotions, personality, and even neurological disorder. This abundant source of data might be used in the complex system to create information in response to generate evidences for identifying the particular person, some of which was also investigated in our previous work (1). This paper is the extension of human gait recognition that presents the analysis of trade-off between evolution of Genetic Programs (GPs) and their performance, considering different training cases, provided that the computational effort and some other parameter are kept constant. Furthermore, our previous work (2) had left an important question unanswered about how the extension of number of fitness cases and the use of experts in collaborative filtering affects the evolution of GPs and recognition of Gait. This study is an attempt to explore the same unexplained question.

저하된 로봇 비전에서의 물체 인식을 위한 진화적 생성 기반의 컬러 검출 기법

This paper introduces GP(Genetic Programming) based color detection model for an object detection of humanoid robot vision. Existing color detection methods have used linear/nonlinear transformation of RGB color-model. However, most of cases have difficulties to classify colors satisfactory because of interference of among color channels and susceptibility for illumination variation. Especially, they are outstanding in degraded images from robot vision. To solve these problems, we propose illumination robust and non-parametric multi-colors detection model using evolution of GP. The proposed method is compared to the existing color-models for various environments in robot vision for real humanoid Nao.

Modeling of Arsenic (III) Removal by Evolutionary Genetic Programming and Least Square Support Vector Machine Models

In this study, the co-precipitation method was used to synthesize the cerium oxide tetraethylenepentamine (CTEPA) hybrid material with the variation of the molar concentration of the metal oxide. Physicochemical techniques, like FESEM, XRD, FTIR, HR-TEM and TGA-DSC, were used to characterize the hybrid material. The adsorption experiment was carried out to estimate the optimum condition for adsorption with the variation of adsorbent dose, pH of the solution, time and initial concentration of the adsorbate. These variables are further used to develop an approach for the evaluation of As(III) removal from water by using evolutionary genetic programming techniques (GP) and least square support vector models (LS-SVM). GP model was found to be the best performing model for understanding the nonlinear behavior and prediction of As(III) removal (minimum standard error GPtraining 0.411, GPtesting 0.658). The adsorption process in this study followed second order kinetics. The experimental data were best fitted to linearly transformed Langmuir isotherm with an R2 (correlation coefficient) value of >0.99 and having a maximum adsorption capacity of 124.8 mg/g at 25 ± 2 °C. The type of adsorption process was derived from the Dubinin–Radushkevich isotherm model. A comparison between the model and actual experimental values gave a high correlation coefficient value (R2 training (GP) 0.988, R2 testing (GP) 0.977).

HBC-Evo: predicting human breast cancer by exploiting amino acid sequence-based feature spaces and evolutionary ensemble system.

We developed genetic programming (GP)-based evolutionary ensemble system for the early diagnosis, prognosis and prediction of human breast cancer. This system has effectively exploited the diversity in feature and decision spaces. First, individual learners are trained in different feature spaces using physicochemical properties of protein amino acids. Their predictions are then stacked to develop the best solution during GP evolution process. Finally, results for HBC-Evo system are obtained with optimal threshold, which is computed using particle swarm optimization. Our novel approach has demonstrated promising results compared to state of the art approaches.

Speed Control of Brushless DC Motor based on Fractional Order PID Controller

Increasing the methods of order calculus for Fractional Order Proportional Integral Derivative (FOPID) controller leads to a wide applications for this type of controller in control systems. A closed loop speed control for BrushLess Direct Current (BLDC) motor with FOPID controller runs the motor very close to the reference speed, provides a good performance and robustness compared with a corresponding system using conventional PID controller. In this paper, the BLDC motor is modeled and simulated in Matlab/Simulink for speed control strategy. A DC link speed control strategy is proposed and implemented. The FOPID speed controller parameters (Kp, ki, kd, λ ) are optimized by genetic evolutionary programming module. The effectiveness of the proposed speed control is verified through simulation. The control system is extra tested under a conditions of some motor parameters perturbations. The simulation results reveal a flexible and stable control strategy.

Fitness approximation for bot evolution in genetic programming

Estimating the fitness value of individuals in an evolutionary algorithm in order to reduce the computational expense of actually calculating the fitness has been a classical pursuit of practitioners. One area which could benefit from progress in this endeavour is bot evolution, i.e. the evolution of non-playing characters in computer games. Because assigning a fitness value to a bot (or rather, the decision tree that controls its behaviour) requires playing the game, the process is very costly. In this work, we introduce two major contributions to speed up this process in the computer game Unreal Tournament 2004™. Firstly, a method for fitness value approximation in genetic programming which is based on the idea that individuals that behave in a similar fashion will have a similar fitness. Thus, similarity of individuals is taken at the performance level, in contrast to commonly employed approaches which are either based on similarity of genotypes or, less frequently, phenotypes. The approximation performs a weighted average of the fitness values of a number of individuals, attaching a confidence level which is based on similarity estimation. The latter is the second contribution of this work, namely a method for estimating the similarity between individuals. This involves carrying out a number of tests consisting of playing a `static' version of the game (with fixed inputs) with the individuals whose similarity is under evaluation and comparing the results. Because the tests involve a limited version of the game, the computational expense of the similarity estimation plus that of the fitness approximation is much lower than that of directly calculating the fitness. The success of the fitness approximation by similarity estimation method for bot evolution in UT2K4 allows us to expect similar results in environments that share the same characteristics.

Exploring Spectroscopic Regression Modelling Using Eureqa: Black-Box Chemometrics or a Useful Tool for Exploration and Learning?

“Eureqa (pronounced ‘eureka’) is a software tool for detecting equations and hidden mathematical relationships in your data” made available by the Cornell Creative Machines Lab. ( http://creativemachines.cornell.edu/eureqfa ). Based on the use of evolutionary genetic programming, the program is capable of testing a wide variety of mathematical functions, and combinations thereof, to find relationships between data. Functions include the standard functions used in multi-linear regression (MLR; +, –, /, *, k) plus others, such as exponentials, roots, trigonometric etc. The objective of this investigation was to determine if this program might be useful in the investigation and development of spectroscopic calibrations. Two sets (173 or 241 samples) of forages (hays) and by-products (hulls, stalks etc.), some of which had been chemically treated with sodium chlorite to increase digestibility, were studied using Eureqa. Data for six analytes were available, but based on previous work using partial least squares regression, only crude protein, a well-determined analyte, and lignin, a much poorly determined analyte, were examined. Results indicated that standard spectral pre-treatments such as normalisation, mean centring, variance scaling, multiplicative scatter correction and derivatives might be beneficial in more rapidly obtaining the best calibrations, but were not necessary unless needed for data scaling. While overall results were comparable to those from partial least squares, but never as quite as good, several aspects of the program led to the conclusion that it is a wonderful exploration and learning tool for spectroscopy, even if one is not interested in developing MLR-based calibrations. As just one example, the results for all samples are displayed for each equation as developed and any of the best dozen or so can be displayed with a simple mouse click. Thus, it is easy to see how the developing equation effects the predicted versus actual fit, outliers etc. in nearly real time. Similarly, the many function fitting options (different error measures, both least- and non-least squares) allow one to explore and see how these affect both the final results and equation evolution. In conclusion, Eureqa is an exceedingly interesting and useful program, both for developing MLR-based calibrations and for studies of calibration development in general.

Genetic Evolutionary Approach for Cutting Forces Prediction in Hard Milling

Hard milling is a very common used machining procedure in the last years. Therefore the prediction of cutting forces is important. The paper deals with this prediction using genetic evolutionary programming (GEP) approach to set mathematical expression for out cutting forces. In this study, face milling was performed using DIN1.2842 (90MnCrV8) cold work tool steel, with a hardness of 61 HRC. Experimental parameters were selected using stability measurements and simulations. In the hard milling experiments, cutting force data in a total of three axes were collected. Feed direction (Fx ) and tangential direction (Fy ) cutting forces generated using genetic evolutionary programming were modelled. Cutting speed and feed rate values were treated as inputs in the models, and average cutting force values as output. Mathematical expressions were created to predict average Fx and Fy forces that can be generated in hard material milling.

Prediction of Layer Thickness in Molten Borax Bath with Genetic Evolutionary Programming

In this study, the vanadium carbide coating in molten borax bath process is modeled by evolutionary genetic programming (GEP) with bath composition (borax percentage, ferro vanadium (Fe-V) percentage, boric acid percentage), bath temperature, immersion time, and layer thickness data. Five inputs and one output data exist in the model. The percentage of borax, Fe-V, and boric acid, temperature, and immersion time parameters are used as input data and the layer thickness value is used as output data. For selected bath components, immersion time, and temperature variables, the layer thicknesses are derived from the mathematical expression. The results of the mathematical expressions are compared to that of experimental data; it is determined that the derived mathematical expression has an accuracy of 89%.

Prediction of Layer Thickness in Molten Borax Bath with Genetic Evolutionary Programming

Prediction of Layer Thickness in Molten Borax Bath with Genetic Evolutionary Programming

밝기변화에 강인한 Genetic Programming 기반의 비파라미터 다중 컬러 검출 모델

본 논문은 물체인식이나 영상추적에 사용되는 컬러검출을 위한 GP(Genetic Programming) 기반의 컬러검출 모델을 제안한다. 기존의 컬러검출은 기본적인 RGB 모델에 대한 선형, 비선형 함수의 변환을 사용하거나, 최적화 기법이나 학습기법에 의해 조명 변화에 개선된 컬러 모델을 사용하고 있다. 하지만 대부분의 경우 색상 채널간의 간섭에 의해 다양한 색상에 대한 분류가 어렵고, 조명변화에 강인하지 못하다. 본 연구에서는 GP의 최적화된 학습기법과 모델 생성 기법을 통해 조명변화에 강인하고, 다중의 색상 검출이 가능하며, 파라미터 설정이 필요 없는 컬러 모델을 제안한다. 제안된 방법을 다양한 색상과 조명환경이 다른 영상에 대해서 기존 컬러모델과 비교 분석하였다. This paper introduces GP(Genetic Programming) based color detection model for an object detection and tracking. Existing color detection methods have used linear/nonlinear transformatin of RGB color-model and improved color model for illumination variation by optimization or learning techniques. However, most of cases have difficulties to classify various of colors because of interference of among color channels and are not robust for illumination variation. To solve these problems, we propose illumination robust and non-parametric multi-colors detection model using evolution of GP. The proposed method is compared to the existing color-models for various colors and images with different lighting conditions.

Parsimonious genetic programming for complex process intelligent modeling: algorithm and applications

A novel genetic programming (GP) algorithm called parsimonious genetic programming (PGP) for complex process intelligent modeling was proposed. First, the method uses traditional GP to generate nonlinear input–output model sets that are represented in a binary tree structure according to special decomposition method. Then, it applies orthogonal least squares algorithm (OLS) to estimate the contribution of the branches, which refers to basic function term that cannot be decomposed anymore, to the accuracy of the model, so as to eliminate complex redundant subtrees and enhance convergence speed. Finally, it obtains simple, reliable and exact linear in parameters nonlinear model via GP evolution. Simulations validate that the proposed method can generate more robust and interpretable models, which is obvious and easy for realization in real applications. For the proposed algorithm, the whole modeling process is fully automatic, which is a rather promising method for complex process intelligent modeling.

Genetic Programming을 이용한 코너 검출자의 자동생성

본 논문은 영상처리에 사용되는 코너점 추출을 위한 GP(Genetic Programming)기반의 코너 검출자를 소개한다. Harris와 SUSAN등 기존의 대표적인 코너 검출자들이 소개되어 왔고, 여러 가지 경험적인 알고리즘들이 연산 시간과 정확도 측면에서 이들 기법을 개선하기 위해서 연구되어 오고 있다. 이들 기법들은 코너점에 대한 특성을 고찰하여 이를 알고리즘화한 것으로 효율성이 높으나, 한편으로 기존의 방식이나 알고리즘에서 크게 벗어난 혁신적인 알고리즘을 발견하기에는 한계가 있다. 본 연구에서는 GP의 진화연산에 의해 자동적으로 코너 검출자를 생성함으로서 새로운 기법의 가능성을 발견하고자 한다. 제안된 방법을 다른 코너 검출자들과 테스트영상을 통해 비교 분석 하였다. This paper introduces GP(Genetic Programming) based corner detectors for an image processing. Various empirical algorithms have been studied to improve computational speed and accuracy including typical approaches, such as Harris and SUSAN. The these techniques are highly efficient, because properties of corner points are inspected and reflected into the algorithms. However these approaches are limited in discovering an innovative algorithm. In this study, we try to discover a more efficient technique by creating corner detector automatically using evolution of GP. The proposed method is compared to the existing corner detectors for test images.

Linear-in-Parameter Models Based on Parsimonious Genetic Programming Algorithm and Its Application to Aero-Engine Start Modeling

A novel Parsimonious Genetic Programming (PGP) algorithm together with a novel aero-engine optimum data-driven dynamic start process model based on PGP is proposed. In application of this method, first, the traditional Genetic Programming(GP) is used to generate the nonlinear input-output models that are represented in a binary tree structure; then, the Orthogonal Least Squares algorithm (OLS) is used to estimate the contribution of the branches of the tree (refer to basic function term that cannot be decomposed anymore according to special rule) to the accuracy of the model, which contributes to eliminate complex redundant subtrees and enhance GP's convergence speed; and finally, a simple, reliable and exact linear-in-parameter nonlinear model via GP evolution is obtained. The real aero-engine start process test data simulation and the comparisons with Support Vector Machines (SVM) validate that the proposed method can generate more applicable, interpretable models and achieve comparable, even superior results to SVM.