Öz yinelemeli sinirsel ağların sinirsel atım verilerinden yararlanılarak modellenmesi hakkında

Year 2021, , 54 - 66, 21.12.2021

Özgür Doruk , Mohammed Al-akam

https://doi.org/10.53525/jster.999008

Abstract

Gerçekçi uyaran-tepki verilerinden ateşleme hızına dayalı uyarıcı ve engelleyici sinir ağının kestirimini amaçlayan teorik ve hesaplamalı bir çalışma sunuyoruz. Uyaran ve tepki kayıtları, Diptera sınıfına mensup H1 nöronları üzerinde bir ölçüm gerçekleştiren önceki bir çalışmadan alınmıştır. Parametre tahmini, maksimum olabilirlik yöntemi ile yapılmaktadır. Uyaran-tepki verileri 20 dakikalık tek bir kayıt olduğu için bölümlere ayrılır ve istatistiksel bilgi içeriğini artırmak için ayrı bölümler üst üste bindirilir. Sentetik veri kullanmadığımız için model parametrelerinin gerçek değerleri bilinmemektedir. Bu nedenle, kaydedilen ve model yanıtlarının kesişme aralıklarını karşılaştırmak için iki örnekli Kolmogorov-Smirnov testi uygulanmıştır. Tahmin ve analiz sonuçları tablo ve grafik şeklinde sunulmaktadır. Ayrıca, değiştirilmiş bir Fitzhugh-Nagumo modelinin kullanıldığı önceki araştırmalarla bir karşılaştırma yapılmıştır.

Keywords

Ateşleme hızı tabanlı sinirsel ağlar, uyarıcı-yatıştırıcı sinirsel ağlar, sinirsel atım, Kolmogorov-Smirnov Testi, atım arası zamanları

On modeling of a recurrent neural network from neural spiking data.

Abstract

We present a theoretical and computational work, aiming at the estimation of firing rate based excitatory and inhibitory neural network from realistic stimulus-response data. The stimulus and response recordings are taken from a previous study which performs a measurement on the H1 neurons of the order Diptera flies. The parameter estimation is performed by maximum likelihood method. As the stimulus-response data is a single recording of 20 minutes, it is segmented and individual segments are superimposed on each other to increase the statistical content of information. The true values of the model parameters are unknown as we are not using synthetic data. Because of this fact, two sample Kolmogorov-Smirnov test is applied to compare the interspiking intervals of the recorded and model responses. Estimation and analysis results are presented in tabular and graphical forms. In addition, a comparison with previous research employing a modified Fitzhugh-Nagumo model is made.

Keywords

: Firing rate based neural networks, Excitatory-inhibitory neurons, Neural spiking, Kolmogorov-Smirnov tests, Interspiking intervals

References

• [1] A. L. Hodgkin and A. F. Huxley, “A quantitative description of membrane current and its application to conduction and excitation in nerve,” J Physiol-London, vol. 117, no. 4, p. 500, 1952. [Online]. Available: https://doi.org/10.1113%2Fjphysiol.1952.sp004764
• [2] R. FitzHugh, “Impulses and physiological states in theoretical models of nerve membrane,” Biophys J, vol. 1, no. 6, pp. 445–466, 1961.
• [3] C. Morris and H. Lecar, “Voltage oscillations in the barnacle giant muscle fiber,” Biophys J, vol. 35, no. 1, pp. 193–213, 1981.
• [4] J. L. Hindmarsh and R. Rose, “A model of neuronal bursting using three coupled first order differential equations,” Proc R Soc Lond B, vol. 221, no. 1222, pp. 87–102, 1984.
• [5] V. Booth, J. Rinzel, and O. Kiehn, “Compartmental model of vertebrate motoneurons for ca2+-dependent spiking and plateau potentials under pharmacological treatment,” J Neurophysiol, vol. 78, no. 6, pp. 3371– 3385, 1997.
• [6] R. O. DORUK, “Neuron modeling: estimating the parameters of aneuron model from neural spiking data,” Turkish Journal of Electrical Engineering & Computer Sciences, vol. 26, no. 5, pp. 2301–2314, 2018.
• [7] V. Mante, R. A. Frazor, V. Bonin, W. S. Geisler, and M. Carandini, “Independence of luminance and contrast in natural scenes and in the early visual system,” Nat Neurosci, vol. 8, no. 12, p. 1690, 2005.
• [8] T. Hosoya, S. A. Baccus, and M. Meister, “Dynamic predictive coding by the retina,” Nature, vol. 436, no. 7047, p. 71, 2005.
• [9] N. C. Rust, O. Schwartz, J. A. Movshon, and E. P. Simoncelli, “Spatiotemporal elements of macaque v1 receptive fields,” Neuron, vol. 46, no. 6, pp. 945–956, 2005.
• [10] E. H. Adelson and J. R. Bergen, “Spatiotemporal energy models for the perception of motion,” Josa a, vol. 2, no. 2, pp. 284–299, 1985.
• [11] A. Borst and F. E. Theunissen, “Information theory and neural coding,” Nat Neurosci, vol. 2, no. 11, p. 947, 1999.
• [12] H. Barlow, “Possible principles underlying the transformation of sensory messages. in. w. rosenblith (ed.) sensory communication (pp. 217-234),” 1959.
• [13] A. L. Fairhall, G. D. Lewen, W. Bialek, and R. R. d. R. van Steveninck, “Efficiency and ambiguity in an adaptive neural code,” Nature, vol. 412, no. 6849, p. 787, 2001.
• [14] B. Hassenstein and W. Reichardt, “Systemtheoretische analyse der zeit-reihenfolgen-und vorzeichenauswertung bei der bewegungsperzeption des r¨usselk¨afers chlorophanus,” Z Naturforsch B, vol. 11, no. 9-10, pp. 513–524, 1956.
• [15] O. R. Doruk and K. Zhang, “Adaptive stimulus design for dynamic recurrent neural network models,” Frontiers in neural circuits, vol. 12, p. 119, 2019.
• [16] M. N. Shadlen and W. T. Newsome, “Noise, neural codes and cortical organization,” Curr Opin Neurol, vol. 4, no. 4, pp. 569–579, 1994.
• [17] I. J. Myung, “Tutorial on maximum likelihood estimation,” J Math Psychol, vol. 47, no. 1, pp. 90–100, 2003.
• [18] D. R. Brillinger, “Maximum likelihood analysis of spike trains of interacting nerve cells,” Biol Cybern, vol. 59, no. 3, pp. 189–200, 1988.
• [19] L. Paninski, “Maximum likelihood estimation of cascade point-process neural encoding models,” Network-Comp Neural, vol. 15, no. 4, pp. 243–262, 2004.
• [20] E. Chornoboy, L. Schramm, and A. Karr, “Maximum likelihood identification of neural point process systems,” Biol Cybern, vol. 59, no. 4, pp. 265–275, 1988.
• [21] A. C. Smith and E. N. Brown, “Estimating a state-space model from point process observations,” Neural Comput, vol. 15, no. 5, pp. 965–991, 2003.
• [22] A. V. Herz, T. Gollisch, C. K. Machens, and D. Jaeger, “Modeling singleneuron dynamics and computations: a balance of detail and abstraction,” Science, vol. 314, no. 5796, pp. 80–85, 2006.
• [23] J. Ma and J. Tang, “A review for dynamics in neuron and neuronal network,” Nonlinear Dynamics, vol. 89, no. 3, pp. 1569–1578, 2017.
• [24] D. Linaro, M. Storace, and M. Giugliano, “Accurate and fast simulation of channel noise in conductance-based model neurons by diffusion approximation,” PLoS Comput Biol, vol. 7, no. 3, p. e1001102, 2011.
• [25] J. A. White, J. T. Rubinstein, and A. R. Kay, “Channel noise in neurons,” Trends in neurosciences, vol. 23, no. 3, pp. 131–137, 2000.
• [26] M. Lv, C. Wang, G. Ren, J. Ma, and X. Song, “Model of electrical activity in a neuron under magnetic flow effect,” Nonlinear Dynamics, vol. 85, no. 3, pp. 1479–1490, 2016.
• [27] M. Lv and J. Ma, “Multiple modes of electrical activities in a new neuron model under electromagnetic radiation,” Neurocomputing, vol. 205, pp. 375–381, 2016.
• [28] F. Wu, C. Wang, Y. Xu, and J. Ma, “Model of electrical activity in cardiac tissue under electromagnetic induction,” Scientific reports, vol. 6, no. 1, pp. 1–12, 2016.
• [29] C. DiMattina and K. Zhang, “Adaptive stimulus optimization for sensory systems neuroscience,” Front Neural Circuit, vol. 7, 2013.
• [30] ——, “Active data collection for efficient estimation and comparison of nonlinear neural models,” Neural Comput, vol. 23, no. 9, pp. 2242–2288, 2011.
• [31] ——, “How to modify a neural network gradually without changing its input-output functionality,” Neural Comput, vol. 22, no. 1, pp. 1–47, 2010.
• [32] E. P. Lynch and C. J. Houghton, “Parameter estimation of neuron models using in-vitro and in-vivo electrophysiological data,” Frontiers in neuroinformatics, vol. 9, p. 10, 2015.
• [33] R. O. Doruk and K. Zhang, “Fitting of dynamic recurrent neural network models to sensory stimulus-response data,” J Biol Phys, vol. 44, no. 3, pp. 449–469, jun 2018. [Online]. Available: https://doi.org/10.1007%2Fs10867-018-9501-z
• [34] R. O¨ . DORUK, “Fitting a recurrent dynamical neural network to neural spiking data: tackling the sigmoidal gain function issues,” Turkish Journal of Electrical Engineering & Computer Sciences, vol. 27, no. 2, pp. 903–920, 2019.
• [35] R. O. Doruk and L. Abosharb, “Estimating the parameters of fitzhugh– nagumo neurons from neural spiking data,” Brain sciences, vol. 9, no. 12, p. 364, 2019.
• [36] M. A. Frye and M. H. Dickinson, “Fly flight: a model for the neural control of complex behavior,” Neuron, vol. 32, no. 3, pp. 385–388, 2001.
• [37] G. Lewen, W. Bialek, and R. Steveninck, “Neural coding of naturalistic motion stimuli,” Network: Computation in Neural Systems, vol. 12, no. 3, pp. 317–329, 2001.
• [38] K. D. Miller and F. Fumarola, “Mathematical equivalence of two common forms of firing rate models of neural networks,” Neural Comput, vol. 24, no. 1, pp. 25–31, 2012.
• [39] F. J. Massey Jr, “The kolmogorov-smirnov test for goodness of fit,” Journal of the American statistical Association, vol. 46, no. 253, pp. 68–78, 1951.
• [40] G. Marsaglia, W. W. Tsang, J. Wang et al., “Evaluating kolmogorov’s distribution,” Journal of statistical software, vol. 8, no. 18, pp. 1–4, 2003.
• [41] U. T. Eden, “Point process models for neural spike trains,” Neural Signal Processing: Quantitative Analysis of Neural Activity, pp. 45–51, 2008.
• [42] E. N. Brown, R. Barbieri, V. Ventura, R. E. Kass, and L. M. Frank, “The time-rescaling theorem and its application to neural spike train data analysis,” Neural Comput, vol. 14, no. 2, pp. 325–346, 2002.