宇宙學家經常建立數學模擬模型來研究觀察到的宇宙。隨著計算能力的大幅提升，模擬可以進行更多的細節生成，並減少觀察到的宇宙與模擬結果之間的差異。然而，高保真的模擬運算耗時且給分析帶來不便，特別是當涉及大量的輸入參數組合時。因此，選擇一個能夠滿足分析任務需要的輸入參數組合就成為分析過程的重要部分之一。在這項工作中，我們提出了一個交互式視覺系統。它可以幫助用戶有效地理解大規模宇宙學數據的參數空間，並進一步發現有價值的模擬輸入參數組合。我們的系統利用基於 GAN 的代理模型來重建模擬輸出，而無需運行每個實例都要花費大量時間的原始昂貴模擬模型。我們還提取了基於深度神經網絡的代理模型學習到的信息，以促進參數空間的探索。例如，隱藏層的輸出用於估計輸入參數配置之間的模擬輸出相似性。模擬參數的敏感性是使用反向傳播從代理模型中估計出來的。我們通過多個案例研究證明了我們系統的有效性，包括發現有價值的模擬輸入參數配置和子區域分析。

Cosmologists often built a mathematics simulation model to study the observed universe. With the momentous improvements in computing power, the simulation can conduct more details and alleviate the discrepancy between the observed universe and simulation result. However, the high-fidelity simulation is time-consuming and brings inconvenience to the analysis. Especially the simulation often involves a large number of simulation input parameter combinations. Therefore, selecting a combination of input parameters that can meet the needs of the analysis task has become an important part of the analysis process. In this work, we propose an interactive visual system, which helps users efficiently understand the parameter space of large-scale cosmology data and further discover the valuable simulation input parameter combinations. Our system utilizes the GAN-based surrogate models to reconstruct the simulation outputs without running the original expensive simulation that costs lots of time per instance. We also extract information learned by the deep neural network-based surrogate models to facilitate the parameter space exploration. For example, the outputs of the hidden layer are used to estimate the simulation output similarity among input parameter configurations. The sensitivities of the simulation parameters are estimated from the surrogate models using back-propagation. We demonstrate the effectiveness of our system with multiple case studies, including the discovery of valuable simulation input parameter configuration and sub-region analysis.

[1] M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin,S. Ghemawat, G. Irving, M. Isard, et al. Tensorflow: A system for large-scale machine learning. In12th{USENIX}symposium on operating systems design and implementation ({OSDI}16), pp. 265–283, 2016.
[2] A. S. Almgren, J. B. Bell, M. J. Lijewski, Z. Luki ́c, and E. Van An-del. Nyx: A massively parallel AMR code for computational cosmology. Astrophysical Journal, 765(1), 2013. doi: 10.1088/0004-637X/765/1/39
[3]M. Arjovsky, S. Chintala, and L. Bottou. Wasserstein gan.arXiv preprintarXiv:1701.07875, 2017.
[4] M. J. Asher, B. F. Croke, A. J. Jakeman, and L. J. Peeters. A review of surrogate models and their application to groundwater modeling. Water Resources Research, 51(8):5957–5973, 2015.
[5] A. Biswas, C. M.Biwer, D. J. Walters, J. Ahrens, D.Francom, E. Lawrence, R. L. Sandberg, D. A. Fredenburg, and C. Bolme. An interactive exploration tool for high-dimensional datasets: A shock physics case study. Computing in Science & Engineering, 22(2):44–54, 2018.
[6] S. Bruckner and T. Moller. Result-driven exploration of simulation parameter spaces for visual effects design. IEEE Transactions on Visualization and Computer Graphics, 16(6):1468–1476, 2010. doi: 10.1109/TV.2010.190
[7] B. Can and C. Heavey. A comparison of genetic programming and artificial neural networks in metamodeling of discrete-event simulation models. Computers and Operations Research, 39(2):424–436, 2012. doi: 10.1016/j.cor.2011.05.004
[8] X. Chen, Y. Tian, J. Gao, and T. Zhang. An Antenna Design Method Based on Guassian Process Surrogate Model and Differential EvolutionAlgorithm.2020 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization, NEMO2020, pp. 7–10, 2020. doi: 10.1109/NEMO49486.2020.9343425
[9] A. Ciccazzo, G. Di, and P. Vittorio. Support vector machines for surrogate modeling of electronic circuits. pp. 69–76, 2014. doi: 10.1007/s00521-013-1509-5
[10] A. Ciccazzo, G. D. Pillo, and V. Latorre. A SVM Surrogate Model-Based Method for Parametric Yield Optimization. 35(7):1224–1228, 2016.
[11] M. Crocce, S. Pueblas, and R. Scoccimarro. Transients from initial conditions in cosmological simulations. Monthly Notices of the Royal Astronomical Society, 373(1):369–381, 2006.
[12] M. Davis, G. Efstathiou, C. S. Frenk, and S. D. White. The evolution of largescale structure in a universe dominated by cold dark matter. The Astrophysical Journal, 292:371–394, 1985.
[13] I. Demir, C. Dick, and R. Westermann. Multi-charts for comparative 3Densemble visualization. IEEE Transactions on Visualization and Computer Graphics, 20(12):2694–2703, 2014. doi: 10.1109/TVCG.2014.2346448
[14] H. Fang and M. F. Horstemeyer. Global response approximation with radial basis functions. Engineering optimization, 38(04):407–424, 2006.
[15] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio. Generative adversarial nets. In Advances in neural information processing systems, pp. 2672–2680, 2014.
[16] I. Gulrajani, F. Ahmed, M. Arjovsky, V. Dumoulin, and A. C. Courville. Improved training of wasserstein gans. In Advances in neural information processing systems, pp. 5767–5777, 2017.
[17] L. Guo, S. Ye, J. Han, H. Zheng, H. Gao, D. Z. Chen, J. X. Wang, and C. Wang. SSR-VFD: Spatial Super-Resolution for Vector Field Data Analysis and Visualization. IEEE Pacific Visualization Symposium, 2020-June:71–80, 2020. doi: 10.1109/PacificVis48177.2020.8737
[18] J. Han, J. Tao, and C. Wang. FlowNet: A Deep Learning Framework for Clustering and Selection of Streamlines and Stream Surfaces. IEEE Transactions on Visualization and Computer Graphics, 26(4):1732–1744,2020. doi: 10.1109/TVCG.2018.2880207
[19] J. Han and C. Wang. Tsr-tvd: Temporal super-resolution for time-varying data analysis and visualization. IEEE transactions on visualization and computer graphics, 26(1):205–215, 2019.
[20] J. Han and C. Wang. Ssr-tvd: Spatial super-resolution for time-varying data analysis and visualization. IEEE Transactions on Visualization and Computer Graphics, 2020.
[21] S. Hazarika, H. Li, K.-C. Wang, H.-W. Shen, and C.-S. Chou. Nnva: Neu-ral network assisted visual analysis of yeast cell polarization simulation. IEEE transactions on visualization and computer graphics, 26(1):34–44,2019.
[22] K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016.
[23] W. He, J. Wang, H. Guo, H. W. Shen, and T. Peterka. CECAV-DNN:Collective Ensemble Comparison and Visualization using Deep Neural Networks. Visual Informatics, 4(2):109–121, 2020. doi: 10.1016/j.visinf.2020.04.004
[24] W. He, J. Wang, H. Guo, K.-C. Wang, H.-W. Shen, M. Raj, Y. S. Nashed, and T. Peterka. Insitunet: Deep image synthesis for parameter space exploration of ensemble simulations. IEEE transactions on visualization and computer graphics, 26(1):23–33, 2019.
[25] G. Hinton, N. Srivastava, and K. Swersky. Rmsprop: Divide the gradient by a running average of its recent magnitude. Neural networks for machine learning, Coursera lecture 6e, p. 13, 2012.
[26] S. Hosder, L. Watson, B. Grossman, W. Mason, H. Kim, R. Haftka, and S. Cox. Polynomial Response Surface Approximations for the Mul-tidisciplinary Design Optimization of a High Speed Civil Transport. Optimization and Engineering, 2(4):431–452, 2001. doi: 10.1023/A:1016094522761
[27] P. Jiang, Q. Zhou, and X. Shao. Surrogate-Model-Based Design and Optimization. 2020. doi: 10.1007/978-981-15-0731-17
[28] P. Jiang, Q. Zhou, and X. Shao. Surrogate model-based engineering design and optimization. Springer, 2020.
[29] Y. Jin. Surrogate-assisted evolutionary computation: Recent advances and future challenges. Swarm and Evolutionary Computation, 1(2):61–70,2011. doi: 10.1016/j.swevo.2011.05.001
[30] D. P. Kingma and M. Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013.
[31] J. P. Kleijnen. Statistical tools for simulation practitioners. Marcel Dekker,1987.
[32] J. P. Kleijnen. Response surface methodology for constrained simulation optimization: An overview. Simulation Modelling Practice and Theory,16(1):50–64, 2008. doi: 10.1016/j.simpat.2007.10.001
[33] J. P. Kleijnen. Kriging metamodeling in simulation: A review. European Journal of Operational Research, 192(3):707–716, 2009. doi: 10.1016/j.ejor.2007.10.013
[34] S. Koziel and J. W. Bandler. Microwave device modeling using space-mapping and radial basis functions. IEEE MTT-S International Mi-crowave Symposium Digest, 3:799–802, 2007. doi: 10.1109/MWSYM.2007.380079
[35] K. Kurach, M. Lucic, X. Zhai, M. Michalski, and S. Gelly. The gan landscape: Losses, architectures, regularization, and normalization. 2018.
[36] M. Lucic, K. Kurach, M. Michalski, S. Gelly, and O. Bousquet. Are gans created equal? a large-scale study.arXiv preprint arXiv:1711.10337,2017.
[37] Z. Luki ́c, C. W. Stark, P. Nugent, M. White, A. A. Meiksin, and A. Alm-gren. The lyman α forest in optically thin hydrodynamical simulations. Monthly Notices of the Royal Astronomical Society, 446(4):3697–3724,2015.
[38] L. McInnes, J. Healy, and J. Melville. Umap: Uniform manifold approximation and projection for dimension reduction.arXiv preprintarXiv:1802.03426, 2018.
[39] E. Meeds and M. Welling. GPS-ABC: Gaussian process surrogate ap-proximate Bayesian computation. Uncertainty in Artificial Intelligence -Proceedings of the 30th Conference, UAI 2014, pp. 593–602, 2014.
[40]M. Mirza and S. Osindero. Conditional generative adversarial nets.arXivpreprint arXiv:1411.1784, 2014.
[41] T. Miyato, T. Kataoka, M. Koyama, and Y. Yoshida. Spectral normalization for generative adversarial networks.arXiv preprint arXiv:1802.05957,2018.
[42] A. M ̈uller. Integral probability metrics and their generating classes of functions. Advances in Applied Probability, pp. 429–443, 1997.
[43] T. Nguyen and J. Schutt-Aine. Gaussian Process surrogate model for variability analysis of RF circuits. IEEE Electrical Design of Advanced Packaging and Systems Symposium, 2020-December(6):2020–2022, 2020.doi: 10.1109/EDAPS50281.2020.9312886
[44] J. O ̃norbe, F. Davies, Z. Luki ́c, J. Hennawi, and D. Sorini. Inhomogeneous reionization models in cosmological hydrodynamical simulations. Monthly Notices of the Royal Astronomical Society, 486(3):4075–4097, 2019.
[45] D. Orban, D. F. Keefe, A. Biswas, J. Ahrens, and D. Rogers. Drag and track: A direct manipulation interface for contextualizing data instances within a continuous parameter space. IEEE transactions on visualization and computer graphics, 25(1):256–266, 2018.
[46] A. Radford, L. Metz, and S. Chintala. Unsupervised representation learning with deep convolutional generative adversarial networks.arXivpreprint arXiv:1511.06434, 2015.
[47] S. Rahvar. Cosmic initial conditions for a habitable universe. Monthly Notices of the Royal Astronomical Society, 470(3):3095–3102, 2017.
[48] B. Schwabe, J. C. Niemeyer, and J. F. Engels. Simulations of solitoniccore mergers in ultralight axion dark matter cosmologies. Physical ReviewD, 94(4):043513, 2016.
[49] M. Sedlmair, C. Heinzl, S. Bruckner, H. Piringer, and T. M ̈oller. Visual parameter space analysis: A conceptual framework. IEEE Transaction son Visualization and Computer Graphics, 20(12):2161–2170, 2014.
[50] M. Tyan, N. V. Nguyen, and J.-W. Lee. Improving variable-fidelity modelling by exploring global design space and radial basis function networks for a erofoil design. Engineering Optimization, 47(7):885–908, 2015.
[51] D. Ustun, M. Tekbas, and A. Toktas. Determination of feed point by surrogate model based on radial basis function for rectangular microstripantennas.2019 International Conference on Artificial Intelligence and Data Processing Symposium, IDAP 2019, (1):6–8, 2019. doi: 10.1109/IDAP.2019.8875956
[52] F. A. Viana, T. W. Simpson, V. Balabanov, and V. Toropov. Special section on multidisciplinary design optimization: metamodeling in multidisciplinary design optimization: how far have we really come? AIAA journal,52(4):670–690, 201
[53] Bruno Villasenor, Brant Robertson, Piero Madau, and Evan Schneider. Effects of
photoionization and photo heating only a forest properties from cholla cosmological simulations. The Astrophysical Journal, 912(2):138,2021.
[54] Michael Walther, Eric Armengaud, Corentin Ravoux, Nathalie Palanque-Delabrouille, Christophe Y`eche, and Zarija Luki´c. Simulating intergalactic gas for desi-like small scale lymana forest observations. Journal of Cosmology and Astroparticle Physics, 2021(04):059, 2021.
[55] Junpeng Wang, Subhashis Hazarika, ChengLi, and Han Wei Shen. Visualization and visual analysis of ensemble data:Asurvey. IEEE Transactions on Visualization and Computer Graphics, 25(9):2853–2872,2019.
[56] Junpeng Wang, Xiaotong Liu, Han Wei Shen, and Guang Lin. Multi-Resolution Climate Ensemble Parameter Analysis with Nested Parallel Coordinates Plots. IEEE
Transactions on Visualization and Computer Graphics, 23(1):81–90,2017.
[57] Mi Xiao, Liang Gao, Xinyu Shao, Haobo Qiu, and PingJiang. A generalized collaborative optimization method and its combination with kriging metamodels for engineering design. Journal of Engineering Design, 23(5):379–399,2012.
[58] Wojciech Zaremba, Ilya Sutskever, and Oriol Vinyals. Recurrent neural network regularization. arXiv preprintarXiv:1409.2329, 2014.
[59] Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei AEfros. Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE international conference on computer vision, pages2223–2232,2017.