Evolutionary Machine Learning: A Survey

Evolutionary Computation (EC) approaches are inspired by nature and solve optimization problems in a stochastic manner. They can offer a reliable and effective approach to address complex problems in real-world applications. EC algorithms have recently been used to improve the performance of Machine Learning (ML) models and the quality of their results. Evolutionary approaches can be used in all three parts of ML: preprocessing (e.g., feature selection and resampling), learning (e.g., parameter setting, membership functions, and neural network topology), and postprocessing (e.g., rule optimization, decision tree/support vectors pruning, and ensemble learning). This article investigates the role of EC algorithms in solving different ML challenges. We do not provide a comprehensive review of evolutionary ML approaches here; instead, we discuss how EC algorithms can contribute to ML by addressing conventional challenges of the artificial intelligence and ML communities. We look at the contributions of EC to ML in nine sub-fields: feature selection, resampling, classifiers, neural networks, reinforcement learning, clustering, association rule mining, and ensemble methods. For each category, we discuss evolutionary machine learning in terms of three aspects: problem formulation, search mechanisms, and fitness value computation. We also consider open issues and challenges that should be addressed in future work.

A high accuracy modeling scheme for dynamic systems: spacecraft reaction wheel model

Reaction wheels are crucial actuators in spacecraft attitude control subsystem (ACS). The precise modeling of reaction wheels is of fundamental need in spacecraft ACS for design, analysis, simulation, and fault diagnosis applications. The complex nature of the reaction wheel leads to modeling difficulties utilizing the conventional modeling schemes. Additionally, the absence of reaction wheel providers’ parameters is crucial for triggering a new modeling scheme. The Radial Basis Function Neural Network (RBFNN) has an efficient architecture, alluring generalization properties, invulnerability against noise, and amazing training capabilities. This research proposes a promising modeling scheme for the spacecraft reaction wheel utilizing RBFNN and an improved variant of the Quantum Behaved Particle Swarm Optimization (QPSO). The problem of enhancing the network parameters of the RBFNN at the training phase is formed as a nonlinear constrained optimization problem. Thus, it is proposed to efficiently resolve utilizing an enhanced version of QPSO with mutation strategy (EQPSO-2M). The proposed technique is compared with the conventional QPSO algorithm and different variants of PSO algorithms. Evaluation criteria rely upon convergence speed, mean best fitness value, stability, and the number of successful runs that has been utilized to assess the proposed approach. A non-parametric test is utilized to decide the critical contrast between the results of the proposed algorithm compared with different algorithms. The simulation results demonstrated that the training of the proposed RBFNN-based reaction wheel model with enhanced parameters by EQPSO-2M algorithm furnishes a superior prediction accuracy went with effective network architecture.

Optimized Production of Biodiesel Using Internet of Things Sensed Temperature with Hybrid Particle Swarm Optimization

Biofuel production offers a non-fossil fuel that can be utilized in modern engines without any redesign. Regardless of receiving rising attention, many researchers have explored microalgae-based biofuel production and found biodiesel production is cost-effective compared to petroleum-centered conventional fuels. The primary reason is that the lipid accumulation of microalgae is possible. An efficient technique is proposed for optimized biodiesel manufacturing with microalgae through an IoT device with the hybrid particle swarm optimization (HPSO) algorithm for elapsing such drawbacks. First, the component of biodiesel is determined. Then, from the components, the temperature value is sensed through the IoT device. Based on the obtained temperature, the reaction parameters are optimized with HPSO to increase productivity and reduce cost. Finally, we observed performance and comparative analysis. The experimental results contrasted with the existent particle swarm optimization (PSO) and genetic algorithm (GA) concerning iteration’s temperature, concentration, production, and fitness. The present HPSO algorithm has differed from the existing PSO and GA concerning IoT sensed temperature and production function. Fitness value and instance concentration are the performance parameters. It varies based on the iteration values. Thus, the proposed optimized biodiesel production is advanced when weighed down with the top-notch methods.

Storage assignment optimization for fishbone robotic mobile fulfillment systems

Robotic Mobile Fulfillment System (RMFS) affects the traditional scheduling problems heavily while operating a warehouse. This paper focuses on storage assignment optimization for Fishbone Robotic Mobile Fulfilment Systems (FRMFS). Based on analyzing operation characteristics of FRMFS, a storage assignment optimization model is proposed with the objectives of maximizing operation efficiency and balancing aisle workload. Adaptive Genetic Algorithm (AGA) is designed to solve the proposed model. To validate the effectiveness of AGA in terms of iteration and optimization rate, this paper designs a variety of scenarios with different task sizes and storage cells. AGA outperforms other four algorithm in terms of fitness value and convergence and has better convergence rate and stability. The experimental results also show the advancement of AGA in large size FRMFS. In conclusion, this paper proposes a storage assignment model for FRMFS to reduce goods movement and travel distance and improve the order picking efficiency.

Boosted Weighted Optimized Convolutional Neural Network Ensemble Classification for Lung Cancer Prediction

One of the major causes of cancer-related mortality worldwide is lung tumors. An earlier prediction of lung tumors is crucial since it may severely increase the death rates. For this reason, genomic profiles have been considered in many advanced microarray technology schemes. Amongst, an Improved Dragonfly optimization Algorithm (IDA) with Boosted Weighted Optimized Neural Network Ensemble Classification (BWONNEC) has been developed which extracts most suitable features and fine-tunes the weights related to the ensemble neural network classifiers. But, its major limitations are the number of learning factors in neural network and computational difficulty. Therefore in this article, a Boosted Weighted Optimized Convolutional Neural Network Ensemble Classification (BWOCNNEC) algorithm is proposed to lessen the number of learning factors and computation cost of neural network. In this algorithm, the boosting weights are combined into the CNN depending on the least square fitness value. Then, the novel weight values are assigned to the features extracted by the IDA. Moreover, these weight values and the chosen features are processed in different CNN structures within the boosted classifier. Further, the best CNN structure in each iteration i.e., CNNs having the least weighted loss is selected and ensemble to predict and diagnose the lung tumors effectively. Finally, the investigational outcomes exhibit that the IDA-BWOCNNEC achieves better prediction efficiency than the existing algorithms.

A meta-parameter tuning model to improve the genetic algorithms design of labeling diversity mappers

Evolutionary algorithms (EAs) have recently been applied to Uncoded Space-Time Labeling Diversity (USTLD) systems to produce labeling diversity mappers. However, the most challenging task is choosing the best parameter setting for the EA to create a more ‘optimal’ mapper design. This paper proposes a ‘meta-Genetic Algorithm (GA)’ used to tune hyperparameters for the Labeling Diversity EA. The algorithm is examined on 16, 32 and 64QAM; 32 and 64PSK; 16, 32 and 64APSK and 16APSK constellations that do not show diagonal symmetry. Furthermore, the meta-GA settings and original GA settings are compared in terms of the number of generations taken to converge to a solution. For QAM constellations, the output using the meta-GA settings matched but did not improve with the original settings. However, the number of generations needed to converge to a solution took 120 times less than the number of generations using the original settings. In the 64PSK constellation, a diversity gain of [Formula: see text][Formula: see text]dB was observed while improving on the actual fitness value from 0.0575 to 0.0661. Similarly, with 32APSK constellation, an improvement in fitness value from 0.1457 to 0.1748 was made while showing diversity gains of [Formula: see text][Formula: see text]dB. 64APSK constellation fitness value improved from 0.0708 to 0.0957, and a [Formula: see text][Formula: see text]dB gain was observed. The most significant improvement was made by the asymmetric 16APSK constellation, with gains of [Formula: see text][Formula: see text]dB and increasing its fitness value three times (0.0981 to 0.3000). A study of the effects of optimizing the GA parameters shows that the number of swaps during crossover [Formula: see text] and the radius [Formula: see text] were the two most important variables to optimize when executing this GA.

Extended Jaccard Indexive Buffalo Optimized Clustering on Geo-social Networks with Big Data

Clustering is a general task of data mining where partitioning a large dataset into dissimilar groups is done. The enormous growth of Geo-Social Networks (GeoSNs) includes users, who create millions of heterogeneous data with a variety of information. Analyzing such volume of data is a challenging task. The clustering of large volume of data is used to identify the frequently visited location information of the users in Geo-Social Networks. In order to improve the clustering of a large volume of data, a novel technique called Extended Jaccard Indexive Buffalo Optimized Data Clustering (EJIBODC) is introduced for grouping the data with high accuracy and less time consumption. The main aim of EJIBODC technique is to partition the big dataset into different groups. In this technique, many clusters with centroids are initialized to group the data. After that, Extended Jaccard Indexive Buffalo Optimization technique is applied to find the fittest cluster for grouping the data. The Extended Jaccard Index is applied in the Buffalo Optimization to measure the fitness between the data and the centroid. Based on the similarity value, using a gradient ascent function, the data finds the fittest cluster centroid for grouping. After that, the fitness value of cluster is updated and all the data gets grouped into a suitable cluster with high accuracy and minimum error rate. An experimental procedure is involved with big geo-social dataset and testing of different clustering algorithms. The series discussion is carried out on factors such as clustering accuracy, error rate, clustering time and space complexity with respect to a number of data. Experimental outcomes demonstrate that the proposed EJIBODC technique achieves improved performance in terms of higher clustering accuracy, less error rate, time consumption and space complexity when compared to previous related clustering techniques.

Pengaruh Penambahan Latihan Core Stability Terhadap Peningkatan Nilai Kesegaran Jasmani B Pada Prajurit

An Army soldier must be able to foster physical fitness so that his body becomes healthy and can carry out his main duties to maintain the sovereignty and territorial integrity of the Republic of Indonesia. This study aims to determine the effect of adding core stability exercise to increasing the value of physical fitness B in soldiers. Quasi experimental design research held at Yon Armed 4 Cimahi. A sample of 20 soldiers divided into 2 groups. The dependent variable is the physical fitness score B and the independent variable is the core stability exercise. Exercise is carried out for 4 weeks with a frequency of 3 times per week. The difference in test results between the experimental group and the control group both before and after core stability exercise was tested by independent T test. There was a difference in the mean physical fitness scores between the experimental group (Mean= 85.80; SD= 2.82) and the control group (Mean= 82.50; SD= 2.32) before the intervention, and the difference was statistically significant (p= 0.011). There was a difference in physical fitness scores before and after the intervention between the experimental group (Mean= 3.90; SD= 1.52) and the control group (Mean= 1.60; SD= 0.70) and statistically significant (p < 0.001), with effect size = 1.94. Conclusion core stability exercise is effective to increase the physical fitness value of B on soldiers of Yon Armed 4 Cimahi

Short-Term Traffic Flow Forecasting Model Based on GA-TCN

Traffic flow forecasting is the key to an intelligent transportation system (ITS). Currently, the short-term traffic flow forecasting methods based on deep learning need to be further improved in terms of accuracy and computational efficiency. Therefore, a short-term traffic flow forecasting model GA-TCN based on genetic algorithm (GA) optimized time convolutional neural network (TCN) is proposed in this paper. The prediction error was considered as the fitness value and the genetic algorithm was used to optimize the filters, kernel size, batch size, and dilations hyperparameters of the temporal convolutional neural network to determine the optimal fitness prediction model. Finally, the model was tested using the public dataset PEMS. The results showed that the average absolute error of the proposed GA-TCN decreased by 34.09%, 22.42%, and 26.33% compared with LSTM, GRU, and TCN in working days, while the average absolute error of the GA-TCN decreased by 24.42%, 2.33%, and 3.92% in weekend days, respectively. The results indicate that the model proposed in this paper has a better adaptability and higher prediction accuracy in short-term traffic flow forecasting compared with the existing models. The proposed model can provide important support for the formulation of a dynamic traffic control scheme.

Loss prediction of three-level amplified spontaneous emission sources in radiation environment

A model of three-level amplified spontaneous emission (ASE) sources, considering radiation effect, is proposed to predict radiation induced loss of output power in radiation environment. Radiation absorption parameters of ASE sources model are obtained by the fitting of color centers generation and recovery process of and gain loss data at lower dose rate. Gain loss data at higher dose is applied for self-validating. This model takes both the influence of erbium ions absorption and photon bleaching effect into consideration, which makes the prediction of different dose and dose rate more accurate and flexible. The fitness value between ASE model and gain loss data is 99.98%, which also satisfies the extrapolation at the low dose rate. The method and model may serve as a valuable tool to predict ASE performance in harsh environment.