{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,9]],"date-time":"2026-06-09T03:00:45Z","timestamp":1780974045298,"version":"3.54.1"},"reference-count":63,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Expert Systems with Applications"],"published-print":{"date-parts":[[2026,7]]},"DOI":"10.1016\/j.eswa.2026.131953","type":"journal-article","created":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T17:14:34Z","timestamp":1772730874000},"page":"131953","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["MishReLU: A hybrid activation function for deep neural networks"],"prefix":"10.1016","volume":"319","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-2289-9593","authenticated-orcid":false,"given":"Naseem","family":"Al-Nawi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8865-9755","authenticated-orcid":false,"given":"Alia","family":"Madain","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5200-1959","authenticated-orcid":false,"given":"Malak","family":"Abdullah","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.eswa.2026.131953_bib0001","series-title":"2024 IEEE symposium on wireless technology applications (ISWTA)","first-page":"173","article-title":"Revolutionizing image recognition: Next-generation CNN architectures for handwritten digits and objects","author":"Absur","year":"2024"},{"issue":"5","key":"10.1016\/j.eswa.2026.131953_bib0002","doi-asserted-by":"crossref","first-page":"717","DOI":"10.1016\/S0731-7085(99)00272-1","article-title":"Basic concepts of artificial neural network (ANN) modeling and its application in pharmaceutical research","volume":"22","author":"Agatonovic-Kustrin","year":"2000","journal-title":"Journal of Pharmaceutical and Biomedical Analysis"},{"key":"10.1016\/j.eswa.2026.131953_bib0003","first-page":"1493","article-title":"Adaptive multi-column deep neural networks with application to robust image denoising","volume":"26","author":"Agostinelli","year":"2013","journal-title":"Advances in Neural Information Processing Systems"},{"issue":"4","key":"10.1016\/j.eswa.2026.131953_bib0004","doi-asserted-by":"crossref","DOI":"10.3390\/agriengineering7040118","article-title":"Hybrid CNN-LSTM model with custom activation and loss functions for predicting fan actuator states in smart greenhouses","volume":"7","author":"Airlangga","year":"2025","journal-title":"AgriEngineering"},{"key":"10.1016\/j.eswa.2026.131953_bib0005","doi-asserted-by":"crossref","first-page":"82249","DOI":"10.1109\/ACCESS.2021.3085855","article-title":"Improving the performance of deep neural networks using two proposed activation functions","volume":"9","author":"Alkhouly","year":"2021","journal-title":"IEEE Access"},{"issue":"22","key":"10.1016\/j.eswa.2026.131953_bib0006","doi-asserted-by":"crossref","first-page":"61789","DOI":"10.1007\/s11042-023-15766-3","article-title":"A novel optimized parametric hyperbolic tangent swish activation function for 1d-CNN: application of sensor-based human activity recognition and anomaly detection","volume":"83","author":"Ankalaki","year":"2024","journal-title":"Multimedia Tools and Applications"},{"key":"10.1016\/j.eswa.2026.131953_bib0007","series-title":"Calculus, 10th edition","author":"Anton","year":"2012"},{"key":"10.1016\/j.eswa.2026.131953_bib0008","series-title":"Pattern recognition and machine learning","author":"Bishop","year":"2006"},{"key":"10.1016\/j.eswa.2026.131953_bib0009","series-title":"Astrophysics source code library (ASCL)","first-page":"1806","article-title":"Keras: The python deep learning library","author":"Chollet","year":"2018"},{"key":"10.1016\/j.eswa.2026.131953_bib0010","unstructured":"Clevert, D. A., Unterthiner, T., & Hochreiter, S. (2015). Fast and accurate deep network learning by exponential linear units (ELUs)."},{"issue":"6","key":"10.1016\/j.eswa.2026.131953_bib0011","doi-asserted-by":"crossref","first-page":"141","DOI":"10.1109\/MSP.2012.2211477","article-title":"The MNIST database of handwritten digit images for machine learning research [best of the web]","volume":"29","author":"Deng","year":"2012","journal-title":"IEEE Signal Processing Magazine"},{"key":"10.1016\/j.eswa.2026.131953_bib0012","doi-asserted-by":"crossref","unstructured":"Dubey, S. R., Singh, S. K., & Chaudhuri, B. B. (2022). Activation functions in deep learning: A comprehensive survey and benchmark. arXiv: 2109.14545.","DOI":"10.1016\/j.neucom.2022.06.111"},{"issue":"1","key":"10.1016\/j.eswa.2026.131953_bib0013","first-page":"163","article-title":"Survey of neural transfer functions","volume":"2","author":"Duch","year":"1999","journal-title":"Neural Computing Surveys"},{"issue":"1","key":"10.1016\/j.eswa.2026.131953_bib0014","doi-asserted-by":"crossref","first-page":"292","DOI":"10.1007\/s11227-023-05441-7","article-title":"The effect of activation functions on accuracy, convergence speed, and misclassification confidence in CNN text classification: a comprehensive exploration","volume":"80","author":"Emanuel","year":"2024","journal-title":"The Journal of Supercomputing"},{"key":"10.1016\/j.eswa.2026.131953_bib0015","series-title":"Deep learning","author":"Goodfellow","year":"2016"},{"key":"10.1016\/j.eswa.2026.131953_bib0016","unstructured":"Google (2020). Tensorboard | tensorflow. https:\/\/www.tensorflow.org\/tensorboard. Accessed: 2020-08-03."},{"key":"10.1016\/j.eswa.2026.131953_bib0017","doi-asserted-by":"crossref","unstructured":"Goyal, M., Goyal, R., Reddy, P. V., & Lall, B. (2019). Activation functions. Springer International Publishing, In Deep Learning: Algorithms and Applications (pp. 1\u201330), 865, 10.1007\/978-3-030-31760-7_1.","DOI":"10.1007\/978-3-030-31760-7_1"},{"key":"10.1016\/j.eswa.2026.131953_bib0018","unstructured":"Hasan, M. M., Hossain, M. A., Srizon, A. Y., & Sayeed, A. (2023). TaLU: A hybrid activation function combining tanh and rectified linear unit to enhance neural networks. arXiv: 2305.04402."},{"key":"10.1016\/j.eswa.2026.131953_bib0019","series-title":"2016 IEEE conference on computer vision and pattern recognition (CVPR)","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"10.1016\/j.eswa.2026.131953_bib0020","series-title":"Deep learning","author":"Heaton","year":"2016"},{"key":"10.1016\/j.eswa.2026.131953_bib0021","series-title":"Applied data science in tourism: Interdisciplinary approaches, methodologies, and applications","first-page":"231","article-title":"Hyperparameter tuning","author":"Ippolito","year":"2022"},{"key":"10.1016\/j.eswa.2026.131953_bib0022","series-title":"Icassp 2025 - 2025 ieee international conference on acoustics, speech and signal processing (icassp)","first-page":"1","article-title":"Enhancing out-of-distribution detection through dynamic activation function","author":"Ji","year":"2025"},{"key":"10.1016\/j.eswa.2026.131953_bib0023","unstructured":"Kaggle Inc. (2017). Explore and run machine learning code with kaggle notebooks. https:\/\/www.kaggle.com\/kernels."},{"issue":"3","key":"10.1016\/j.eswa.2026.131953_bib0024","doi-asserted-by":"crossref","first-page":"286","DOI":"10.3103\/S1060992X24700152","article-title":"Enhancement of neural network performance with the use of two novel activation functions: modexp and modexpm","volume":"33","author":"Kalim","year":"2024","journal-title":"Optical Memory and Neural Networks"},{"key":"10.1016\/j.eswa.2026.131953_bib0025","article-title":"Rsigelu: A nonlinear activation function for deep neural networks","volume":"9","author":"Kili\u00e7arslan","year":"2021","journal-title":"Expert Systems with Applications"},{"issue":"11","key":"10.1016\/j.eswa.2026.131953_bib0026","doi-asserted-by":"crossref","first-page":"15485","DOI":"10.1007\/s11227-024-06057-1","article-title":"Trish: An efficient activation function for CNN models and analysis of its effectiveness with optimizers in diagnosing glaucoma","volume":"80","author":"K\u00f6zkurt","year":"2024","journal-title":"The Journal of Supercomputing"},{"key":"10.1016\/j.eswa.2026.131953_bib0027","series-title":"Technical Report","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"10.1016\/j.eswa.2026.131953_bib0028","series-title":"Efficient backprop","first-page":"9","author":"LeCun","year":"2002"},{"key":"10.1016\/j.eswa.2026.131953_bib0029","doi-asserted-by":"crossref","DOI":"10.1016\/j.neunet.2025.107375","article-title":"Cauchy activation function and XNet","volume":"188","author":"Li","year":"2025","journal-title":"Neural Networks"},{"key":"10.1016\/j.eswa.2026.131953_bib0030","doi-asserted-by":"crossref","first-page":"23722","DOI":"10.1109\/ACCESS.2018.2817593","article-title":"Breast cancer classification based on fully-connected layer first convolutional neural networks","volume":"6","author":"Liu","year":"2018","journal-title":"IEEE Access"},{"issue":"9","key":"10.1016\/j.eswa.2026.131953_bib0031","doi-asserted-by":"crossref","first-page":"1273","DOI":"10.1109\/TC.2018.2817237","article-title":"A stochastic computational multi-layer perceptron with backward propagation","volume":"67","author":"Liu","year":"2018","journal-title":"IEEE Transactions on Computers"},{"key":"10.1016\/j.eswa.2026.131953_bib0032","series-title":"Proceedings of the 2019 IEEE 5th international conference on computer and communications (ICCC), Chengdu, China","first-page":"2000","article-title":"Natural-logarithm-rectified activation function in convolutional neural networks","author":"Liu","year":"2019"},{"key":"10.1016\/j.eswa.2026.131953_bib0033","series-title":"Proceedings of the 30th international conference on machine learning (ICML)","first-page":"1","article-title":"Rectifier nonlinearities improve neural network acoustic models","author":"Maas","year":"2013"},{"issue":"3","key":"10.1016\/j.eswa.2026.131953_bib0034","doi-asserted-by":"crossref","DOI":"10.3390\/axioms12030246","article-title":"Nipuna: A novel optimizer activation function for deep neural networks","volume":"12","author":"Madhu","year":"2023","journal-title":"Axioms"},{"key":"10.1016\/j.eswa.2026.131953_bib0035","doi-asserted-by":"crossref","unstructured":"Misra, D. (2020). Mish: A self regularized non-monotonic activation function. arXiv: 1908.08681.","DOI":"10.5244\/C.34.191"},{"issue":"20","key":"10.1016\/j.eswa.2026.131953_bib0036","doi-asserted-by":"crossref","DOI":"10.3390\/math12203174","article-title":"State-of-the-art results with the fashion-MNIST dataset","volume":"12","author":"Mukhamediev","year":"2024","journal-title":"Mathematics"},{"key":"10.1016\/j.eswa.2026.131953_bib0037","first-page":"5324","article-title":"SERF Towards better training of deep neural networks using log-softplus ERror activation function","volume":"47","author":"Nag","year":"2023","journal-title":"IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)"},{"key":"10.1016\/j.eswa.2026.131953_bib0038","series-title":"2020 international joint conference on neural networks (IJCNN)","first-page":"1","article-title":"Improving the performance of neural networks with an ensemble of activation functions","author":"Nandi","year":"2020"},{"key":"10.1016\/j.eswa.2026.131953_bib0039","unstructured":"Nwankpa, C., Ijomah, W., Gachagan, A., & Marshall, S. (2018). Activation functions: Comparison of trends in practice and research for deep learning. arXiv: 1811.03378."},{"key":"10.1016\/j.eswa.2026.131953_bib0040","unstructured":"Ramachandran, P., Zoph, B., & Le, Q. V. (2017). Searching for activation functions. arXiv: 1710.05941."},{"key":"10.1016\/j.eswa.2026.131953_bib0041","series-title":"2020 IEEE 18th world symposium on applied machine intelligence and informatics (SAMI)","first-page":"281","article-title":"A review of activation function for artificial neural network","author":"Rasamoelina","year":"2020"},{"issue":"6088","key":"10.1016\/j.eswa.2026.131953_bib0042","doi-asserted-by":"crossref","first-page":"533","DOI":"10.1038\/323533a0","article-title":"Learning representations by back-propagating errors","volume":"323","author":"Rumelhart","year":"1986","journal-title":"Nature"},{"key":"10.1016\/j.eswa.2026.131953_bib0043","doi-asserted-by":"crossref","first-page":"85","DOI":"10.1016\/j.neunet.2014.09.003","article-title":"Deep learning in neural networks: An overview","volume":"61","author":"Schmidhuber","year":"2015","journal-title":"Neural Networks"},{"issue":"9","key":"10.1016\/j.eswa.2026.131953_bib0044","doi-asserted-by":"crossref","DOI":"10.3390\/app12094140","article-title":"Arabic language opinion mining based on long short-term memory (LSTM)","volume":"12","author":"Setyanto","year":"2022","journal-title":"Applied Sciences"},{"issue":"12","key":"10.1016\/j.eswa.2026.131953_bib0045","first-page":"310","article-title":"Activation functions in neural networks","volume":"6","author":"Sharma","year":"2017","journal-title":"Towards Data Sci"},{"key":"10.1016\/j.eswa.2026.131953_bib0046","unstructured":"Subramanian, B., Jeyaraj, R., Ugli, R. A. A., & Kim, J. (2024). Apalu: A trainable, adaptive activation function for deep learning networks. arXiv: 2402.08244."},{"key":"10.1016\/j.eswa.2026.131953_bib0047","series-title":"Review and comparison of commonly used activation functions for deep neural networks","first-page":"203","author":"Szanda\u0142a","year":"2021"},{"key":"10.1016\/j.eswa.2026.131953_bib0048","doi-asserted-by":"crossref","unstructured":"Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2014). Going deeper with convolutions. newblock In Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR) (pp. 1\u20139) arXiv: 1409.4842.","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"10.1016\/j.eswa.2026.131953_bib0049","unstructured":"Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR) (pp. 4700\u20134708). arXiv: 1608.06993."},{"key":"10.1016\/j.eswa.2026.131953_bib0050","doi-asserted-by":"crossref","first-page":"168626","DOI":"10.1109\/ACCESS.2024.3474574","article-title":"Parametric activation functions for neural networks: A tutorial survey","volume":"12","author":"S\u00e1ra Pusztah\u00e1zi","year":"2024","journal-title":"IEEE Access"},{"key":"10.1016\/j.eswa.2026.131953_bib0051","unstructured":"K. Team (2020a). Keras documentation: IMDB movie review sentiment classification dataset. Accessed on November 28, 2020https:\/\/keras.io\/api\/datasets\/imdb\/."},{"key":"10.1016\/j.eswa.2026.131953_bib0052","unstructured":"K. Team (2020b). Keras documentation: Reuters newswire classification dataset. Accessed on September 2, 2020https:\/\/keras.io\/api\/datasets\/reuters\/."},{"key":"10.1016\/j.eswa.2026.131953_bib0053","series-title":"Proceedings of human language technologies: The 2009 annual conference of the north american chapter of the association for computational linguistics","first-page":"245","article-title":"Quadratic features and deep architectures for chunking","volume":"vol. Companion Volume","author":"Turian","year":"2009"},{"key":"10.1016\/j.eswa.2026.131953_bib0054","doi-asserted-by":"crossref","first-page":"88","DOI":"10.1016\/j.neucom.2019.07.017","article-title":"ReLTanh: An activation function with vanishing gradient resistance for SAE-based DNNs and its application to rotating machinery fault diagnosis","volume":"363","author":"Wang","year":"2019","journal-title":"Neurocomputing"},{"issue":"4","key":"10.1016\/j.eswa.2026.131953_bib0055","doi-asserted-by":"crossref","DOI":"10.3390\/electronics11040540","article-title":"Smish: A novel activation function for deep learning methods","volume":"11","author":"Wang","year":"2022","journal-title":"Electronics"},{"key":"10.1016\/j.eswa.2026.131953_bib0056","series-title":"2021 7th international conference on computer and communications (ICCC)","first-page":"1444","article-title":"Mmrelu: A simple and smooth activation function with high convergence speed","author":"Wu","year":"2021"},{"key":"10.1016\/j.eswa.2026.131953_bib0057","unstructured":"Xiao, H., Rasul, K., & Vollgraf, R. (2017). Fashion-MNIST: a novel image dataset for benchmarking machine learning algorithms. arXiv: 1708.07747."},{"key":"10.1016\/j.eswa.2026.131953_bib0058","unstructured":"Xinyu, L. (2020). Tanhexp: A smooth activation function with high convergence speed for lightweight neural networks. arXiv: 2006.05792."},{"issue":"1","key":"10.1016\/j.eswa.2026.131953_bib0059","doi-asserted-by":"crossref","first-page":"288","DOI":"10.1038\/s41377-022-00976-5","article-title":"Reconfigurable nonlinear photonic activation function for photonic neural network based on non-volatile opto-resistive RAM switch","volume":"11","author":"Xu","year":"2022","journal-title":"Light: Science & Applications"},{"key":"10.1016\/j.eswa.2026.131953_bib0060","doi-asserted-by":"crossref","first-page":"72727","DOI":"10.1109\/ACCESS.2020.2987829","article-title":"Rmaf: Relu-memristor-like activation function for deep learning","volume":"8","author":"Yu","year":"2020","journal-title":"IEEE Access"},{"key":"10.1016\/j.eswa.2026.131953_bib0061","unstructured":"Zahedi, L., Mohammadi, F. G., Rezapour, S., Ohland, M. W., & Amini, M. H. (2021). Search algorithms for automated hyper-parameter tuning. arXiv: 2104.14677."},{"key":"10.1016\/j.eswa.2026.131953_bib0062","series-title":"European conference on computer vision (ECCV)","first-page":"818","article-title":"Visualizing and understanding convolutional networks","author":"Zeiler","year":"2014"},{"key":"10.1016\/j.eswa.2026.131953_bib0063","unstructured":"Zhang, G., & Li, H. (2018). Effectiveness of scaled exponentially-regularized linear units (SERLUs). arXiv: 1807.10117."}],"container-title":["Expert Systems with Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0957417426008663?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0957417426008663?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,6,9]],"date-time":"2026-06-09T02:55:39Z","timestamp":1780973739000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0957417426008663"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7]]},"references-count":63,"alternative-id":["S0957417426008663"],"URL":"https:\/\/doi.org\/10.1016\/j.eswa.2026.131953","relation":{},"ISSN":["0957-4174"],"issn-type":[{"value":"0957-4174","type":"print"}],"subject":[],"published":{"date-parts":[[2026,7]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"MishReLU: A hybrid activation function for deep neural networks","name":"articletitle","label":"Article Title"},{"value":"Expert Systems with Applications","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.eswa.2026.131953","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}],"article-number":"131953"}}