Activation functions are the most significant properties of artificial neural networks (ANN) because these functions are directly related with the ability of ANN in learning or modelling a system or a function. Furthermore, another reason for the significance of the fact that determination of optimal activation function in ANN is its relationship with success level. In this experimental study, the effects of different types of wire electrodes, cooling techniques and workpiece materials on surface roughness (Ra) and cutting speed (Vc) in wire electrical discharge machining (WEDM) were investigated by using trainable activation functions (AFt) and modelling them in ANNs. So far, a number of methods have been performed according to the data set in order to optimally predict Ra and Vc results. Among these methods, randomized ANN with AFt was found to be the best one for robust prediction according to RMSE values. While the value was 0.280 for Vc, it was 0.2104 for Ra. Optimum activation functions in Ra and Vc were found at first and third degree trainable functions, respectively.
LinCTChungIFHuangSY.Improvement of machining accuracy by fuzzy logic at corner parts for wire-EDM. Fuzzy Sets Syst2001;
122: 499–511.
2.
SharmaPChakradharDNarendranathS.Effect of wire diameter on surface integrity of wire electrical discharge machined Inconel 706 for gas turbine application. J Manuf Process2016;
24: 170–178.
3.
ManjaiahMNarendranathSBasavarajappaS, et al.
Effect of electrode material in wire electro discharge machining characteristics of Ti50Ni50-xCux shape memory alloy. Precis Eng2015;
41: 68–77.
4.
NewtonTRMelkoteSNWatkinsTR, et al.
Investigation of the effect of process parameters on the formation and characteristics of recast layer in wire-EDM of Inconel 718. Mater Sci Eng A2009;
513–514: 208–215.
5.
GürbüzHBadayŞHamaratİ.Effects of different wire electrodes and cooling methods on surface roughness in wire electro discharge machining. Afyon Kocatepe Univ J Sci Eng2018;
18: 676–683.
6.
GürbüzHBadayŞ. The effects of wire electrode diameter, coated and uncoated wire electrodes on cutting speed in CNC wire electro discharge machining. In: 8th international advanced technologies symposium (IATS’17), Elazığ, Turkey, 2017, pp. 978–984. Fırat Universty.
7.
KuriakoseSShunmugamMS.Characteristics of wire-electro discharge machined Ti6Al4V surface. Mater Lett2004;
58: 2231–2237.
8.
SahaAMondalSC.Multi-objective optimization in WEDM process of nanostructured hard facing materials through hybrid techniques. Measurement2016;
94: 46–59.
9.
PrasadNBVParameswaraRao ChVS and MalleswaraRaoS.Studies on wire selection for machining with WEDM. Int J Comput Sci Trends Technol2015;
3: 300–303.
10.
KhanAAAliMBMShaffiarNBM.Relationship of surface roughness with current and voltage during wire EDM. J Appl Sci2006;
6: 2317–2320.
11.
KlockeFLungDThomaidisD, et al.
Using ultra-thin electrodes to produce micro-parts with wire-EDM. J Mater Process Technol2004;
149: 579–584.
12.
StrakaLCornyIPitelJ.Properties evaluation of thin microhardened surface layer of tool steel after wire EDM. Metals2016;
6: 1–16.
13.
MajumderA.Comparison of ANN with RSM in predicting surface roughness with respect to process parameters in Nd:Yag laser drilling. Int J Eng Sci Technol2010;
2: 5175–5186.
14.
BadayŞ.Modelling of main cutting forces with artificial neural network in turning of spheroidizing heat treatmented AISI 1050 steel. Technol Appl Sci2016;
11: 1–9.
15.
GürbüzHKurtAŞekerU.Investigation of the effects of different chip breaker forms on the cutting forces using artificial neural networks. Gazi Univ J Sci2012;
25: 803–814.
16.
GürbüzHSözenAŞekerU.Modelling of effects of various chip breaker forms on surface roughness in turning operations by utilizing artificial neural networks. J Polytech2016;
19: 71–83.
17.
UgrasenaGRavindraHVPrakashGVN, et al.
Process optimization and estimation of machining performances using artificial neural network in wire EDM. Proc Mater Sci2014;
6: 1752–1760.
18.
CondeAArriandiagaASanchezJA, et al.
High-accuracy wire electrical discharge machining using artificial neural networks and optimization techniques. Robot Comput Integr Manuf2018;
49: 24–38.
19.
ChenHCLinJCYangYK, et al.
Optimization of wire electrical discharge machining for pure tungsten using a neural network integrated simulated annealing approach. Exp Syst Appl2010;
37: 7147–7153.
20.
SuryaVRKumarKMVKeshavamurthyR, et al.
Prediction of machining characteristics using artificial neural network in wire EDM of Al7075 based in-situ composite. Mater Today Proc2017;
4: 203–212.
21.
SunYZengWMaX, et al.
A hybrid approach for processing parameters optimization of Ti-22Al-25Nb alloy during hot deformation using artificial neural network and genetic algorithm. Intermetallics2011;
19: 1014–1019.
22.
MajumderHMaityK.Application of GRNN and multivariate hybrid approach to predict and optimize WEDM responses for Ni-Ti shape memory alloy. Appl Soft Comput2018;
70: 665–679.
23.
SinghBMisraJP.Surface finish analysis of wire electric discharge machined specimens by RSM and ANN modeling. Measurement2019;
137: 225–237.
24.
SarkarSGhoshKMitraS, et al.
An integrated approach to optimization of WEDM combining single-pass and multipass cutting operation. Mater Manuf Process2010;
25: 799–807.
25.
RamakrishnanaRKarunamoorthybL.Modeling and multi-response optimization of Inconel 718 on machining of CNC WEDM process. J Mater Process Technol2008;
207: 343–349.
26.
ThankachanTPrakashKSMaliniR, et al.
Prediction of surface roughness and material removal rate in wire electrical discharge machining on aluminum based alloys/composites using Taguchi coupled grey relational analysis and artificial neural networks. Appl Surf Sci2019;
472: 22–35.
27.
SchmidtWFKraaijveld MA and DuinRPW. Feed forward neural networks with random weights. In: 11th international conference on pattern recognition, Hague, Netherlands, 30 August-3 September 1992, pp. 1–4. Piscataway: IEEE.
28.
ErtuğrulÖF.A novel type of activation function in artificial neural networks: trained activation function.Neural Netw2018;
99: 148–157.
29.
NaseemITogneriRBennamounM.Linear regression for face recognition.IEEE Trans Pattern Anal Mach Intell2010;
32: 2106–2112.
30.
PerlichCSwirszczG.On cross-validation and stacking: building seemingly predictive models on random data. SIGKDD Explor2011;
12: 11–15.
31.
ArlotSCelisseA.A survey of cross-validation procedures for model selection.Stat Surv2010;
4: 40–79.
32.
ProhaszkaJMamalisAGVaxevanidisNM.The effect of electrode material on machinability in wire electro-discharge machining. J Mater Process Technol1997;
69: 233–237.
33.
ChalisgaonkarRKumarJ.Investigation of the machining parameters and integrity of the work and wire surfaces after finish cut WEDM of commercially pure titanium. J Braz Soc Mech Sci Eng2016;
38: 883–911.
34.
RasitA.Artificial neural networks applications in wind energy systems: a review. Renew Sustain Energy Rev2015;
49: 534–562.
35.
BorchaniHVarandoGBielzaC, et al.
A survey on multi‐output regression. Wiley Interdiscip Rev Data Min Knowl Discov2015;
5: 216–233.
36.
TanMSongXYangX, et al.
Support-vector-regression machine technology for total organic carbon content prediction from wireline logs in organic shale: a comparative study. J Nat Gas Sci Eng2015;
26: 792–802.
37.
ErtuğrulÖFTağlukME.A novel version of k nearest neighbor: dependent nearest neighbor. Appl Soft Comput2017;
55: 480–490.
38.
RahmanMAHossenJVenkataseshaiahCH, et al.
A survey of machine learning techniques for self-tuning Hadoop performance.Int J Electr Comput Eng2018;
8: 1854–1861.
39.
JiangJChenCMaJ, et al.
SRLSP: a face image super-resolution algorithm using smooth regression with local structure prior. IEEE Trans Multimed2017;
19: 27–40.
40.
ChengJHSunDW.Partial least squares regression (PLSR) applied to NIR and HSI spectral data modeling to predict chemical properties of fish muscle. Food Eng Rev2017;
9: 36–49.
41.
RichardsonRROsborneMAHoweyDA.Gaussian process regression for forecasting battery state of health. J Power Sources2017;
357: 209–219.
