Skip to main content

Advertisement

SpringerLink
Log in

Heuristic techniques to optimize neural network architecture in manufacturing applications

  • Original Article
  • Published:
Neural Computing and Applications Aims and scope Submit manuscript

Abstract

Nowadays application of neural networks in the manufacturing field is widely assessed even if this type of problem is typically characterized by an insufficient availability of data for a robust network training. Satisfactory results can be found in the literature, in both forming and machining operations, regarding the use of a neural network as a predictive tool. Nevertheless, the research of the optimal network configuration is still based on trial-and-error approaches, rather than on the application of specific techniques . As a consequence, the best method to determine the optimal neural network configuration is still a lack of knowledge in the literature overview. According to that, a comparative analysis is proposed in this work. More in detail four different approaches have been used to increase the generalization abilities of a neural network. These methods are based, respectively, on the use of genetic algorithms, Taguchi, tabu search and decision trees. The parameters taken into account in this work are the training algorithm, the number of hidden layers, the number of neurons and the activation function of each hidden layer. These techniques have been firstly tested on three different datasets, generated through numerical simulations in the Deform2D environment, in an attempt to map the input–output relationship for an extrusion, a rolling and a shearing process. Subsequently, the same approach has been validated on a fourth dataset derived from the literature review for a complex industrial process to widely generalize and asses the proposed methodology in the whole manufacturing field. Four tests were carried out for each dataset modifying the original data with a random noise with zero mean and standard deviation of one, two and five per cent. The results show that the use of a suitable technique for determining the architecture of a neural network can generate a significant performance improvement compared to a trial-and-error approach.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

References

  1. Neves J, Rocha M, Cortez P (2007) Evolution of neural networks for classification and regression. Neurocomputing 70(16):2809–2816

    Google Scholar 

  2. Bouzerdoum A, Arulampalam G (2003) A generalized feedforward neural network architecture for classification and regression. Neural Netw 16(5):561–568

    Google Scholar 

  3. Pfingsten JT (2007) Machine learning for mass production and industrial engineering

  4. Pospichal J, Svozil D, Kvasnicka V (1997) Introduction to multi-layer feed-forward neural networks. Chemom Intell Lab Syst 39(1):43–62

    Article  Google Scholar 

  5. Umbrello D, Ambrogio G, Filice L, Shivpuri R (2007) An ann approach for predicting subsurface residual stresses and the desired cutting conditions during hard turning. J Mater Process Technol 189(1):143–152

    Article  Google Scholar 

  6. Hambli R (2009) Statistical damage analysis of extrusion processes using finite element method and neural networks simulation. Finite Elem Anal Des 45(10):640–649

    Article  Google Scholar 

  7. Yang F (2010) Neural network metamodeling for cycle time-throughput profiles in manufacturing. Eur J Oper Res 205(1):172–185

    Article  MATH  Google Scholar 

  8. Unal M, Onat M, Demetgul M, Kucuk H (2014) Fault diagnosis of rolling bearings using a genetic algorithm optimized neural network. Measurement 58:187–196

    Article  Google Scholar 

  9. Saravanan R, Venkatesan D, Kannan K (2009) A genetic algorithm-based artificial neural network model for the optimization of machining processes. Neural Comput Appl 18(2):135–140

    Article  Google Scholar 

  10. Yao YL, Fang XD (1993) Assessment of chip forming patterns with tool wear progression in machining via neural networks. Int J Mach Tools Manuf 33(1):89–102

    Article  Google Scholar 

  11. Gagliardi F, Ambrogio G (2013) Design of an optimized procedure to predict opposite performances in porthole die extrusion. Neural Comput Appl 23(1):195–206

    Article  Google Scholar 

  12. Fığlalı A, Özcan B (2014) Artificial neural networks for the cost estimation of stamping dies. Neural Comput Appl 25(3–4):1–10

    Google Scholar 

  13. Yusuff RM, Saberi S (2012) Neural network application in predicting advanced manufacturing technology implementation performance. Neural Comput Appl 21(6):1191–1204

    Article  Google Scholar 

  14. Jerez JM, Gómez I, Franco L (2009) Neural network architecture selection: Can function complexity help? Neural Process Lett 30(2):71–87

    Article  Google Scholar 

  15. Haussler D, Baum EB (1989) What size net gives valid generalization? Neural Comput 1(1):151–160

    Article  Google Scholar 

  16. Yoneyama T, Camargo L (2001) Specification of training sets and the number of hidden neurons for multilayer perceptrons. Neural Comput 13(12):2673–2680

    Article  MATH  Google Scholar 

  17. Qing M, Jun-Hai Z, Xi-Zhao W, Qing-Yan S (2013) Architecture selection for networks trained with extreme learning machine using localized generalization error model. Neurocomputing 102:3–9

    Article  Google Scholar 

  18. Lirov Y (1992) Computer aided neural network engineering. Neural Netw 5(4):711–719

    Article  Google Scholar 

  19. Haykin S (1994) Neural networks: a comprehensive foundation. Prentice Hall, Englewood Cliffs

    MATH  Google Scholar 

  20. Berger JO (1985) Statistical decision theory and Bayesian analysis. Springer, Berlin

    Book  MATH  Google Scholar 

  21. Smith AFM, Bernardo JM (2009) Bayesian theory, vol 405. Wiley, New York

    Google Scholar 

  22. Johnson JD, Sexton RS, Dorsey RE (1999) Optimization of neural networks: a comparative analysis of the genetic algorithm and simulated annealing. Eur J Oper Res 114(3):589–601

    Article  MATH  Google Scholar 

  23. Araújo R, Soares SA, Carlos H (2013) Comparison of a genetic algorithm and simulated annealing for automatic neural network ensemble development. Neurocomputing 121:498–511

    Article  Google Scholar 

  24. Bogart C, Whitley D, Starkweather T (1990) Genetic algorithms and neural networks: optimizing connections and connectivity. Parallel Comput 14(3):347–361

    Article  Google Scholar 

  25. Spears WM, De Jong KA (1992) A formal analysis of the role of multi-point crossover in genetic algorithms. Ann Math Artif Intell 5(1):1–26

    Article  MATH  Google Scholar 

  26. Agrawal RB, Agrawal RB, Deb K (1994) Simulated binary crossover for continuous search space. Complex Syst 9(3):1–15

    MathSciNet  MATH  Google Scholar 

  27. Glover F (1990) Artificial intelligence, heuristic frameworks and tabu search. Manag Dec Econ 11(5):365–375

    Article  Google Scholar 

  28. Li M, Ruan X, Ye J, Qiao J (2007) A tabu based neural network learning algorithm. Neurocomputing 70(4):875–882

  29. Lim LEN, Khaw JFC, Lim BS (1995) Optimal design of neural networks using the Taguchi method. Neurocomputing 7(3):225–245

    Article  MATH  Google Scholar 

  30. Rowlands H, Packianather MS, Drake PR (2000) Optimizing the parameters of multilayered feedforward neural networks through Taguchi design of experiments. Qual Reliab Eng Int 16(6):461–473

    Article  Google Scholar 

  31. Tannock J, Sukthomya W (2005) The optimisation of neural network parameters using Taguchi’s design of experiments approach: an application in manufacturing process modelling. Neural Comput Appl 14(4):337–344

    Article  Google Scholar 

  32. Quinlan JR (1996) Learning decision tree classifiers. ACM Comput Surv (CSUR) 28(1):71–72

    Article  Google Scholar 

  33. Neville PG (1999) Decision trees for predictive modeling. SAS Institute Inc

  34. Kim YS (2008) Comparison of the decision tree, artificial neural network, and linear regression methods based on the number and types of independent variables and sample size. Expert Syst Appl 34(2):1227–1234

    Article  Google Scholar 

  35. Kubat M, Ivanova I (1995) Initialization of neural networks by means of decision trees. Knowl Based Syst 8(6):333–344

    Article  Google Scholar 

  36. Maji P (2008) Efficient design of neural network tree using a new splitting criterion. Neurocomputing 71(4):787–800

    Article  Google Scholar 

  37. Pinkus A, Schocken S, Leshno M, Lin VY (1993) Multilayer feedforward networks with a nonpolynomial activation function can approximate any function. Neural Netw 6(6):861–867

    Article  Google Scholar 

  38. Cybenko G (1989) Approximation by superpositions of a sigmoidal function. Math Control Signals Syst 2(4):303–314

    Article  MathSciNet  MATH  Google Scholar 

  39. Orr GB Müller KR LeCun YA, Bottou L (1998) Efficient backprop. In: Neural networks: tricks of the trade. Springer, Berlin, pp 9–50

  40. Sandberg IW, Park J (1991) Universal approximation using radial-basis-function networks. Neural Comput 3(2):246–257

    Article  Google Scholar 

  41. Argyros AA, Lourakis MIA (2005) Is Levenberg–Marquardt the most efficient optimization algorithm for implementing bundle adjustment? In: Tenth IEEE international conference on computer vision, 2005 (ICCV 2005), vol 2. IEEE, pp 1526–1531

  42. Stafylopatis A, Likas A (2000) Training the random neural network using quasi-newton methods. Eur J Oper Res 126(2):331–339

    Article  MathSciNet  MATH  Google Scholar 

  43. Braun H, Riedmiller M (1993) A direct adaptive method for faster backpropagation learning: The rprop algorithm. In: IEEE international conference on neural networks, 1993. IEEE, pp 586–591

  44. Al Timemy AH, Al Naima FM (2010) Resilient back propagation algorithm for breast biopsy classification based on artificial neural networks. Breast Cancer 4(5):6–7

    Google Scholar 

  45. Khoshayanda MR, Soltani Bozchalooic I, Dadgara A, Rafiee-Tehrania MGhaffari A, Abdollahib H (2006) Performance comparison of neural network training algorithms in modeling of bimodal drug delivery. Int J Pharm 327(1):126–138

  46. Matlab Users Guide. Version 7.13. 0.564 (r2011b). the mathworks, 2011

  47. Gagliardi F, Giardini C, Ceretti E, Fratini L (2009) A new approach to study material bonding in extrusion porthole dies. CIRP Ann Manuf Technol 58(1):259–262

    Article  Google Scholar 

  48. Umbrello D Ambrogio G, Filice L (2004) Numerical analysis of the fracture surface in thick sheets blanking. In: Proceedings of the 7th ESAFORM conference, pp 757–760

  49. Filice L, Ambrogio G, Gagliardi F (2004) First experimental and numerical evidences concerning the hydropiercing process. In: 7th Esaform conference on material forming, Trondheim (Norway), 28–30 April 2004

  50. Fluhrer J (2005) Deform 3d version 5.0 user’s manual [z]. Columbus: Scientific Forming Technologles Corporation

  51. Schmid SR, Kalpakjian S (2010) Manufacturing processes for engineering materials, vol 5. Pearson education

  52. Peascoe RA, Watkins TR, Thiele JD, Melkote SN (2000) Effect of cutting edge geometry and workpiece hardness on surface residual stresses in finish hard turning of aisi 2100 steel. J Manuf Sci Eng 122(1):642–649

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Mechanical, Energy and Management Engineering, University of Calabria, 87036, Rende, Italy

    Claudio Ciancio, Giuseppina Ambrogio, Francesco Gagliardi & Roberto Musmanno

Authors
  1. Claudio Ciancio
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Giuseppina Ambrogio
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Francesco Gagliardi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Roberto Musmanno
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Claudio Ciancio.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ciancio, C., Ambrogio, G., Gagliardi, F. et al. Heuristic techniques to optimize neural network architecture in manufacturing applications. Neural Comput & Applic 27, 2001–2015 (2016). https://doi.org/10.1007/s00521-015-1994-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-015-1994-9

Keywords

Search

Navigation