Abstract
The development of universal and high-efficiency optimization algorithms is a very important research direction of neural networks. Stochastic Gradient Decent Momentum(SGDM) is one of the most successful optimization algorithms, and easily fall into local extremes minimum. Inspired by the prominent success of Fractional-order Calculus in automatic control, we proposed a method based on Fractional-Order named Fractional-Order Momentum(FracM). As a natural extension of integral calculus, fractional order calculus inherits almost all the characteristics of integral calculus, and have some memorization and nonlocality. FracM performs fractional-order difference of momentum and gradient in SGDM algorithm. FracM can partially solve the problem of traps in the local minimum point and accelerated the train process. The proposed FracM optimization method can compare with the most advanced SGDM and Adam and other advanced optimization algorithm in terms of classification accuracy. The experiments show that FracM outperforms other optimizers on CIFAR10/100 and textual datasets IMDB with transformer-based models.
Similar content being viewed by others
References
An W, Wang H, Sun Q, Xu J, Dai Q, Zhang L (2018) A pid controller approach for stochastic optimization of deep networks, pp 8522–8531
Bahdanau D, Cho K, Bengio Y (2014) Neural machine translation by jointly learning to align and translate. arXiv:1409.0473
Bao C, Pu Y, Zhang Y (2018) Fractional-order deep backpropagation neural network. Comput Intell Neurosci 2018:1–10
Bottou L (2010) Large-scale machine learning with stochastic gradient descent. In: Proceedings of COMPSTAT’2010. Springer, pp 177–186
Bottou L (2010) Large-scale machine learning with stochastic gradient descent
Bottou L, Bousquet O (2011) The tradeoffs of large scale learning. Adv Neural Inf Process Syst 20:1–8
Dauphin YN, Pascanu R, Gulcehre C, Cho K, Ganguli S, Bengio Y (2014) Identifying and attacking the saddle point problem in high-dimensional non-convex optimization. In: Advances in neural information processing systems, pp 2933–2941
Ding J, Ren X, Luo R, Sun X (2019) An adaptive and momental bound method for stochastic learning. arXiv:1910.12249
Ding Z, Shen Y (2016) Projective synchronization of nonidentical fractional-order neural networks based on sliding mode controller. Neural Netw 76:97–105
Dubey SR, Chakraborty S, Roy SK, Mukherjee S, Singh SK, Chaudhuri BB (2019) diffgrad: an optimization method for convolutional neural networks. IEEE Trans Neural Netw 1–12
Ginsburg B, Castonguay P, Hrinchuk O, Kuchaiev O, Lavrukhin V, Leary R, Li J, Nguyen H, Zhang Y, Cohen JM (2019) Stochastic gradient methods with layer-wise adaptive moments for training of deep networks. arXiv:1905.11286
Glorot X, Bordes A, Bengio Y (2011) Deep sparse rectifier neural networks, vol 15, pp 315–323
Goodfellow I, Pougetabadie J, Mirza M, Xu B, Wardefarley D, Ozair S, Courville A, Bengio Y (2014) Generative adversarial nets 2672–2680
Gupta V, Koren T, Singer Y (2018) Shampoo: preconditioned stochastic tensor optimization, pp 1837–1845
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition, pp 770–778
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proceedings of the IEEE computer society conference on computer vision and pattern recognition 2016-Decem, pp 770–778. https://doi.org/10.1109/CVPR.2016.90arXiv:1512.03385
Heo B, Chun S, Oh SJ, Han D, Yun S, Uh Y, Ha JW (2020) Slowing down the weight norm increase in momentum-based optimizers. arXiv preprint arXiv:2006.08217
Hinton GE, Osindero S, Teh YW (2006) A fast learning algorithm for deep belief nets. Neural Comput 18:1527–1554
Hinton GE, Srivastava N, Krizhevsky A, Sutskever I, Salakhutdinov R (2012) Improving neural networks by preventing co-adaptation of feature detectors. arXiv:1207.0580
Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv:1502.03167
Kan T, Gao Z, Yang C, Jian S (2021) Convolutional neural networks based on fractional-order momentum for parameter training. Neurocomputing 449:85–99
Kaslik E, Sivasundaram S (2012) Nonlinear dynamics and chaos in fractional-order neural networks. Neural Netw 32:245–256
Kingma DP, Ba J (2014) Adam: a method for stochastic optimization. arXiv:1412.6980
Krizhevsky A, Hinton G et al (2009) Learning multiple layers of features from tiny images
LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444
Ling ZH, Kang SY, Zen H, Senior A, Schuster M, Qian XJ, Meng HM, Deng L (2015) Deep learning for acoustic modeling in parametric speech generation: a systematic review of existing techniques and future trends. IEEE Signal Process Mag 32:35–52
Liu L, Jiang H (2019) On the variance of the adaptive learning rate and beyond. arXiv:1908.03265
Loshchilov I, Hutter F (2017) Decoupled weight decay regularization. arXiv:1711.05101
Lucas J, Zemel R Grosse R (2018) Aggregated momentum: stability through passive damping
Luo L, Xiong Y, Liu Y, Sun X (2019) Adaptive gradient methods with dynamic bound of learning rate
Ma J, Yarats D, (2018) Quasi-hyperbolic momentum and adam for deep learning. arXiv:1810.06801
Ma X (2020) Apollo: an adaptive parameter-wise diagonal quasi-newton method for nonconvex stochastic optimization
Mnih V, Heess N, Graves A, Kavukcuoglu K (2014) Recurrent models of visual attention. arXiv:1406.6247
Nakkiran P, Kaplun G, Bansal Y, Yang T, Barak B, Sutskever I (2019) Deep double descent: where bigger models and more data hurt. arXiv:1912.02292
Noda K, Yamaguchi Y, Nakadai K, Okuno HG, Ogata T (2015) Audio-visual speech recognition using deep learning. Appl Intell 42:722–737
Osawa K, Tsuji Y, Ueno Y, Naruse A, Yokota R, Matsuoka S (2018) Second-order optimization method for large mini-batch: training resnet-50 on imagenet in 35 epochs
Qian N (1999) On the momentum term in gradient descent learning algorithms. Neural Netw 12:145–151
Reddi SJ, Kale S, Kumar S (2018) On the convergence of adam and beyond
Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, Huang Z, Karpathy A, Khosla A, Bernstein M et al (2015) Imagenet large scale visual recognition challenge. Int J Comput Vis 115:211–252
Schmidhuber J (2015) Deep learning in neural networks. Neural Netw 61:85–117
Shazeer N, Stern M (2018) Adafactor: adaptive learning rates with sublinear memory cost. 35th International conference on machine learning, ICML 2018 10, pp 7322–7330. arXiv:1804.04235
Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition
Sun R (2019) Optimization for deep learning: theory and algorithms
Sun W, Su F, Wang L (2017) Improving deep neural networks with multi-layer maxout networks and a novel initialization method. Neurocomputing 278:34–40
Sutskever I, Vinyals O, Le QV (2014) Sequence to sequence learning with neural networks. arXiv:1409.3215
Sutton R (1986) Two problems with back propagation and other steepest descent learning procedures for networks. In: Proceedings of the eighth annual conference of the cognitive science society, pp 823–832
Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A (2015) Going deeper with convolutions, pp 1–9
Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser L, Polosukhin I (2017) Attention is all you need, pp 5998–6008
Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser L, Polosukhin I (2017) Attention is all you need. arXiv
Wu G, Lu W, Gao G, Zhao C, Liu J (2016) Regional deep learning model for visual tracking. Neurocomputing 175:310–323
Yang Y, He Y, Wang Y, Wu M (2018) Stability analysis of fractional-order neural networks: an lmi approach. Neurocomputing 285:82–93
Yao Z, Gholami A, Shen S, Keutzer K, Mahoney MW (2020) Adahessian: an adaptive second order optimizer for machine learning
You Y, Li J, Reddi SJ, Hseu J, Kumar S, Bhojanapalli S, Song X, Demmel J, Hsieh C (2019) Large batch optimization for deep learning: training bert in 76 minutes. arXiv:1904.00962
Yu D, Deng L (2011) Deep learning and its applications to signal and information processing. IEEE Signal Process Mag 28:145–154
Zaheer M, Reddi SJ, Sachan DS, Kale S, Kumar S (2018) Adaptive methods for nonconvex optimization 9793–9803
Zeiler MD (2012) Adadelta: an adaptive learning rate method. arXiv:1212.5701
Zhang MR, Lucas J, Hinton GE, Ba J (2019) Lookahead optimizer: k steps forward, 1 step back. arXiv:1907.08610
Zhang S, Yu Y, Yu J (2016) Lmi conditions for global stability of fractional-order neural networks. IEEE Trans Neural Netw Learn Syst 28:2423–2433
Zhuang J, Tang T, Ding Y, Tatikonda SC, Dvornek NC, Papademetris X, Duncan JS (2020) Adabelief optimizer: adapting stepsizes by the belief in observed gradients. In: Advances in neural information processing systems
Zinkevich M (2003) Online convex programming and generalized infinitesimal gradient ascent, pp 928–935
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Yu, Z., Sun, G. & Lv, J. A fractional-order momentum optimization approach of deep neural networks. Neural Comput & Applic 34, 7091–7111 (2022). https://doi.org/10.1007/s00521-021-06765-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-021-06765-2