Synaptic Weights in a Neuromemristive Radioisotope Classifier

Abstract:

A memristor crossbar’s intrinsic device network dynamics can be harnessed to efficiently conduct radionuclide classification tasks by retrofitting the network with a peripheral CMOS-based architecture that has been structurally and functionally optimized for this classification task. However, the hardware implementation of this classification platform is limited by the physical characteristics of the memristor which has a finite number of states. This renders the employment of traditional neural network learning algorithms, where the weights are not limited to defined states, as an excessively complex task. Hence, this paper tests the limitations on weight resolution and its effect in classification precision when implementing a spiking locally competitive learning algorithm. Both linear and nonlinear weight distributions are examined. The algorithm’s local competitiveness is assessed for the specific application of radionuclide identification. The system is tested using spectra data obtained from the United States National Nuclear Data Center as the classification database dictionary. The platform’s accuracy is measured when test signals with 100, 10 and 1 signal-to-noise ratios are assessed. It has been shown that the system is highly effective for classifying radioisotopes with linear weight distribution even with high levels of noise present. A minor classification accuracy improvement was also observed for weight states distributions with a higher density of values in the low conductivity range. Therefore, it is concluded that a memristor-based radionuclide classifier should have at least 4 possible states mapping the algorithm’s synaptic weights.

• Periodical: Engineering Headway (Volume 12)
• Pages: 37-45
• DOI: https://doi.org/10.4028/p-qI5Dkx
• Online since: September 2024
• Authors: Jorge I. Canales-Verdial*
• Keywords: Memristors, Neuromorphic Computing, Radioisotope Classification

References

[1] Backus, J. (1978). Can programming be liberated from the von Neumann style? A functional style and its algebra of programs. Communications of the ACM, 21(8), 613-641.

DOI: 10.1145/359576.359579

[2] Shalf, J. (2020). The future of computing beyond Moore's Law. Philosophical Transactions of the Royal Society A, 378(2166), 20190061.

DOI: 10.1098/rsta.2019.0061

[3] Dennard, R. H., Cai, J., & Kumar, A. (2018). A perspective on today's scaling challenges and possible future directions. In Handbook of Thin Film Deposition (pp.3-18). William Andrew Publishing.

DOI: 10.1016/b978-1-4377-7873-1.00001-2

[4] Kim, M. S., Kim, M. S., Lee, G. J., Sunwoo, S. H., Chang, S., Song, Y. M., & Kim, D. H. (2022). Bio‐inspired artificial vision and neuromorphic image processing devices. Advanced Materials Technologies, 7(2), 2100144.

DOI: 10.1002/admt.202100144

[5] Blouw, P., & Eliasmith, C. (2020, May). Event-driven signal processing with neuromorphic computing systems. In ICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp.8534-8538). IEEE.

DOI: 10.1109/icassp40776.2020.9053043

[6] Aitsam, M., Davies, S., & Di Nuovo, A. (2022). Neuromorphic computing for interactive robotics: a systematic review. Ieee Access, 10, 122261-122279.

DOI: 10.1109/access.2022.3219440

[7] Mitchell, J. P., Bruer, G., Dean, M. E., Plank, J. S., Rose, G. S., & Schuman, C. D. (2017, October). NeoN: Neuromorphic control for autonomous robotic navigation. In 2017 IEEE International Symposium on Robotics and Intelligent Sensors (IRIS) (pp.136-142). IEEE.

DOI: 10.1109/iris.2017.8250111

[8] Canales-Verdial, J. I., Wagner, J. R., Schmucker, L. A., Wetzel, M., Proctor, P., Carson, M., ... & Zarkesh-Ha, P. (2024). Energy-Efficient Neuromorphic Architectures for Nuclear Radiation Detection Applications. Sensors, 24(7), 2144.

DOI: 10.3390/s24072144

[9] Kulkarni, M. S., & Teuscher, C. (2012, July). Memristor-based reservoir computing. In Proceedings of the 2012 IEEE/ACM International Symposium on Nanoscale Architectures (pp.226-232).

DOI: 10.1145/2765491.2765531

[10] Li, Y., Wang, Z., Midya, R., Xia, Q., & Yang, J. J. (2018). Review of memristor devices in neuromorphic computing: materials sciences and device challenges. Journal of Physics D: Applied Physics, 51(50), 503002.

DOI: 10.1088/1361-6463/aade3f

[11] Hong, Q., Zhao, L., & Wang, X. (2019). Novel circuit designs of memristor synapse and neuron. Neurocomputing, 330, 11-16.

DOI: 10.1016/j.neucom.2018.11.043

[12] Danilin, S. N., Shchanikov, S. A., Zuev, A. D., Bordanov, I. A., & Sakulin, A. E. (2019, October). The research of fault tolerance of memristor-based artificial neural networks. In 2019 12th International Conference on Developments in eSystems Engineering (DeSE) (pp.539-544). IEEE.

DOI: 10.1109/dese.2019.00104

[13] Aguirre, F., Sebastian, A., Le Gallo, M., Song, W., Wang, T., Yang, J. J., ... & Lanza, M. (2024). Hardware implementation of memristor-based artificial neural networks. Nature Communications, 15(1), 1974.

[14] Liu, C., Hu, M., Strachan, J. P., & Li, H. (2017, June). Rescuing memristor-based neuromorphic design with high defects. In Proceedings of the 54th Annual Design Automation Conference 2017 (pp.1-6).

DOI: 10.1145/3061639.3062310

[15] Zhou, W., Wen, S., Liu, Y., Liu, L., Liu, X., & Chen, L. (2023). Forgetting memristor based STDP learning circuit for neural networks. Neural Networks, 158, 293-304.

DOI: 10.1016/j.neunet.2022.11.023

[16] Ben Dayan Rubin, D. D., & Fusi, S. (2007). Long memory lifetimes require complex synapses and limited sparseness. Frontiers in computational neuroscience, 1, 117.

DOI: 10.3389/neuro.10.007.2007

[17] Woods, W., Bürger, J., & Teuscher, C. (2015). Synaptic weight states in a locally competitive algorithm for neuromorphic memristive hardware. IEEE Transactions on Nanotechnology, 14(6), 945-953.

DOI: 10.1109/tnano.2015.2449835

[18] Rozell, C. J., Johnson, D. H., Baraniuk, R. G., & Olshausen, B. A. (2008). Sparse coding via thresholding and local competition in neural circuits. Neural computation, 20(10), 2526-2563.

DOI: 10.1162/neco.2008.03-07-486

[19] Li, C., Belkin, D., Li, Y., Yan, P., Hu, M., Ge, N., ... & Xia, Q. (2018). Efficient and self-adaptive in-situ learning in multilayer memristor neural networks. Nature communications, 9(1), 2385.

DOI: 10.1038/s41467-018-04484-2

[20] Yu, Y., Adu, K., Tashi, N., Anokye, P., Wang, X., & Ayidzoe, M. A. (2020). Rmaf: Relu-memristor-like activation function for deep learning. IEEE Access, 8, 72727-72741.

DOI: 10.1109/access.2020.2987829

[21] NNDC | National Nuclear Data Center. (n.d.). Retrieved April 27, 2024, from https://www.nndc.bnl.gov/

[22] Tuli, J. K. (2004). Nuclear wallet cards for radioactive nuclides. Brookhaven National Laboratory, Upton, New York, USA.