Học máy cho việc giảm quy mô: Sử dụng nhiều quần thể song song trong lập trình di truyền

Springer Science and Business Media LLC - Tập 33 - Trang 1497-1533 - 2019
D. A. Sachindra1,2, S. Kanae1
1Department of Civil and Environmental Engineering, School of Environment and Society, Tokyo Institute of Technology, Tokyo, Japan
2Institute for Sustainability and Innovation, College of Engineering and Science, Victoria University, Melbourne, Australia

Tóm tắt

Trong việc triển khai thuật toán GP truyền thống, do các mô hình phát triển trong một deme duy nhất (một môi trường mà trong đó một quần thể mô hình được phát triển), có thể dẫn đến việc sản xuất ra các mô hình không tối ưu với khả năng tổng quát kém do thiếu sự đa dạng của các mô hình. Để giải quyết vấn đề trên, trong nghiên cứu này, tiềm năng của việc phát triển các mô hình song song trong nhiều deme với các thuộc tính di truyền khác nhau (các môi trường không đồng nhất song song) và sự phát triển tiếp tục của một số mô hình tốt nhất được chọn từ mỗi deme trong một deme khác được gọi là deme chính đã được xem xét, liên quan đến việc giảm quy mô dữ liệu khí hậu lớn đến nhiệt độ tối thiểu hàng ngày (Tmin) và nhiệt độ tối đa hàng ngày (Tmax). Đã phát hiện rằng độc lập với chế độ khí hậu (tức là ấm hoặc lạnh) và vị trí địa lý của trạm quan sát, một phần nhỏ của các mô hình tốt nhất (ví dụ 25%) thu được từ thế hệ cuối cùng của mỗi deme riêng lẻ đủ để hình thành một quần thể ban đầu đa dạng cho deme chính. Ngoài ra, độc lập với chế độ khí hậu và vị trí địa lý của trạm quan sát, cả hai mô hình giảm quy mô Tmin và Tmax hàng ngày được phát triển bằng thuật toán lập trình di truyền nhiều quần thể song song (PMPGP) cho thấy khả năng tổng quát tốt hơn so với các mô hình phát triển bằng thuật toán GP deme đơn truyền thống, ngay cả khi lượng thông tin thừa trong dữ liệu của các biến dự đoán là cao. Các mô hình được phát triển cho Tmin và Tmax hàng ngày với thuật toán PMPGP mô phỏng ít các ngoại lệ lớn phi lý hơn so với các mô hình được phát triển bằng thuật toán GP.

Từ khóa

#lập trình di truyền #giảm quy mô #học máy #mô hình khí hậu #đa dạng mô hình

Tài liệu tham khảo

Abbaspour KC, Rouholahnejad E, Vaghefi S, Srinivasan R, Yang H, Kløve B (2015) A continental-scale hydrology and water quality model for Europe: calibration and uncertainty of a high-resolution large-scale SWAT model. J Hydrol 524:733–752 Anandhi A, Srinivas VV, Nanjundiah RS, Kumar DN (2008) Downscaling precipitation to river basin in India for IPCC SRES scenarios using support vector machine. Int J Climatol 28:401–420 Anandhi A, Srinivas VV, Kumar DN, Nanjundiah RS (2009) Role of predictors in downscaling surface temperature to river basin in India for IPCC SRES scenarios using support vector machine. Int J Climatol 29:583–603 Brands S, Gutiérrez JM, Herrera S, Cofiño AS (2012) On the use of reanalysis data for downscaling. J Clim 25:2517–2526 Chen TS, Yu PS, Tang YH (2010) Statistical downscaling of daily precipitation using support vector machines and multivariate analysis. J Hydrol 385:13–22 Chu JL, Yu PS (2010) A study of the impact of climate change on local precipitation using statistical downscaling. J Geophys Res 115:D10105. https://doi.org/10.1029/2009jd012357 Coulibaly P (2004) Downscaling daily extreme temperatures with genetic programming. Geophys Res Lett 31:L16203. https://doi.org/10.1029/2004gl020075 Danandeh Mehr A, Nourani V, Kahya E, Hrnjica B, Sattar AMA, Yaseen ZM (2018) Genetic programming in water resources engineering: a state-of-the-art review. J Hydrol 566:643–667. https://doi.org/10.1016/j.jhydrol.2018.09.043 Devak M, Dhanya CT (2016) Downscaling of precipitation in Mahanadi Basin, India using support vector machine, K-nearest neighbour and hybrid of support vector machine with K-nearest neighbour. In: Raju N (ed) Geostatistical and geospatial approaches for the characterization of natural resources in the environment. Springer, Cham Devak M, Dhanya CT, Gosain AK (2015) Dynamic coupling of support vector machine and K-nearest neighbour for downscaling daily rainfall. J Hydrol 525:286–301 Eden JM, Widmann M, Grawe D, Rast S (2012) Skill, correction, and downscaling of GCM-simulated precipitation. J Clim 25:3970–3984 Enomoto T, Hoskins BJ, Matsuda Y (2003) The formation mechanism of the Bonin high in August. Q J R Meteor Soc 129:157–178 Erhardt RJ, Band LE, Smith RL, Lopes BJ (2015) Statistical downscaling of precipitation on a spatially dependent network using a regional climate model. Stoch Environ Res Risk Assess 29:1835–1849 Fernandez F, Tomassini M, Vanneschi L (2003) An empirical study of multipopulation genetic programming. Genet Program Evol Mach 4:21–51 Fowler HJ, Blenkinsop S, Tebaldi C (2007) Linking climate change modelling to impacts studies: recent advances in downscaling techniques for hydrological modelling. Int J Climatol 27:1547–1578 Galelli S, Castelletti A (2013) Tree-based iterative input variable selection for hydrological modelling. Water Resour Res 49:4295–4310 Ghosh S, Mujumdar PP (2008) Statistical downscaling of GCM simulations to streamflow using relevance vector machine. Adv Water Resour 31:132–146 Goly A, Teegavarapu RSV, Mondal A (2014) Development and evaluation of statistical downscaling models for monthly precipitation. Earth Interact 18:1–28 Grossman MJ, Zaiki M, Nagata R (2015) Interannual and interdecadal variations in typhoon tracks around Japan. Int J Climatol 35:2514–2527 Hammami D, Lee TS, Ouarda TBMJ, Le J (2012) Predictor selection for downscaling GCM data with LASSO. J Geophys Res Atmos. https://doi.org/10.1029/2012jd017864 Huth R (1999) Statistical downscaling in central Europe: evaluation of methods and potential predictors. Clim Res 13:91–101 Japan Meteorological Agency (2018) http://www.data.jma.go.jp/gmd/cpd/longfcst/en/tourist_japan.html. Accessed 10 Aug 2018 Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, Iredell M, Saha S, White G, Woollen J, Zhu Y, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mo KC, Ropelewski C, Wang J, Leetmaa A, Reynolds R, Jenne R, Joseph D (1996) The NCEP/NCAR 40-year reanalysis project. Bull Am Meteor Soc 77:437–471 Klein WH, Walsh JE (1983) A comparison of pointwise screening and empirical orthogonal functions in specifying monthly surface temperature from 700 mb data. Mon Weather Rev 111:669–673 Koukidis EN, Berg AA (2009) Sensitivity of the statistical downscaling model (SDSM) to reanalysis products. Atmos Ocean 47:1–18. https://doi.org/10.3137/ao924.2009 Koza JR (1992) Genetic programming: on the programming of computers by means of natural selection. MIT Press, Cambridge Lanzante JR, Dixon KW, Nath MJ, Whitlock CE, Adams-Smith D (2018) Some pitfalls in statistical downscaling of future climate. Bull Am Meteor Soc 99:791–803 Laprise R (2008) Regional climate modelling. J Comput Phys 227:3641–3666 Liu W, Fu G, Liu C, Song X, Ouyang R (2013a) Projection of future rainfall for the North China plain using two statistical downscaling models and its hydrological implications. Stoch Environ Res Risk Assess 27:1783–1797 Liu Y, Xie L, Morrison JM, Kamykowski D (2013b) Dynamic downscaling of the impact of climate change on the ocean circulation in the Galápagos Archipelago. Adv Meteorol. https://doi.org/10.1155/2013/837432 Lutz K, Jacobeit J, Philipp A, Seubert S, Kunstmann H, Laux P (2012) Comparison and evaluation of statistical downscaling techniques for station-based precipitation in the Middle East. Int J Climatol 32:1579–1595 Manzanas R, Lucero A, Weisheimer A, Gutiérrez JM (2018) Can bias correction and statistical downscaling methods improve the skill of seasonal precipitation forecasts? Clim Dyn 50:1161–1176 Maraun D, Wetterhall F, Ireson AM, Chandler RE, Kendon EJ, Widmann M, Brienen S, Rust HW, Sauter T, Themel M, Venema VKC, Chun KP, Goodess CM, Jones RG, Onof C, Vrac M, Thiele-Eich I (2010) Precipitation downscaling under climate change: recent developments to bridge the gap between dynamical models and the end user. Rev Geophys. https://doi.org/10.1029/2009rg000314 Matsumoto S, Ninomiya K, Yoshizumi S (1971) Characteristic features of Baiu front associated with heavy rainfall. J Meteor Soc Jpn 49:267–281 May R, Maier H, Dandy G, Fernando TG (2008) Non-linear variable selection for artificial neural networks using partial mutual information. Environ Modell Softw 23:1312–1326 May R, Dandy G, Maier H (2011) Review of input variable selection methods for artificial neural networks. In: Suzuki K (ed) Artificial neural network methodological advances and biomedical applications. IntechOpen, London. https://doi.org/10.5772/16004 Miyasaka T, Nakamura H (2005) Structure and formation mechanisms of the northern hemisphere summertime subtropical highs. J Clim 18:5046–5065 Mujumdar PP, Kumar DN (2012) Floods in a changing climate: hydrologic modelling (international hydrology series). Cambridge University Press, Cambridge. https://doi.org/10.1017/cbo9781139088428 Murazaki K, Kurihara K, Sasaki H (2010) Dynamical downscaling of JRA-25 precipitation over Japan using the MRI-regional climate model. SOLA 6:141–144 Ogi M, Tachibana Y, Yamazaki K (2004) The connectivity of the winter North Atlantic Oscillation (NAO) and the summer Okhotsk high. J Meteor Soc Jpn 82:905–913 Parasuraman K, Elshorbagy A, Carey SK (2007) Modelling the dynamics of the evapotranspiration process using genetic programming. Hydrolog Sci J 52:563–578 Parker WS (2016) Reanalyses and observations: what’s the difference? Bull Am Meteor Soc 97:1565–1572 Pearson K (1896) Mathematical contributions to the theory of evolution III regression heredity and panmixia. Philos Trans R Soc S Afr 187:253–318 Pour SH, Harun SB, Shahid S (2014) Genetic programming for the downscaling of extreme rainfall events on the East Coast of Peninsular Malaysia. Atmosphere 5:914–936 Ratner B (2009) The correlation coefficient: its values range between 1+/− 1, or do they? J Target Meas Anal Mark 17:139–142 Sachindra DA, Perera BJC (2016) Statistical downscaling of general circulation model outputs to precipitation accounting for non-stationarities in predictor–predictand relationships. PLoS ONE 11:e0168701. https://doi.org/10.1371/journal.pone.0168701 Sachindra DA, Huang F, Barton A, Perera BJC (2013) Least square support vector and multi-linear regression for statistically downscaling general circulation model outputs to catchment streamflows. Int J Climatol 33:1087–1106 Sachindra DA, Huang F, Barton A, Perera BJC (2014a) Statistical downscaling of general circulation model outputs to precipitation-part 1: calibration and validation. Int J Climatol 34:3264–3281 Sachindra DA, Huang F, Barton A, Perera BJC (2014b) Statistical downscaling of general circulation model outputs to precipitation-part 2: bias-correction and future projections. Int J Climatol 34:3282–3303 Sachindra DA, Huang F, Barton A, Perera BJC (2014c) Statistical downscaling of general circulation model outputs to catchment scale hydroclimatic variables: issues, challenges and possible solutions. J Water Clim Change 5:496–525 Sachindra DA, Ahmed K, Rashid Md Mamunur, Shahid S, Perera BJC (2018a) Statistical downscaling of precipitation using machine learning techniques. Atmos Res 212:240–258 Sachindra DA, Ahmed K, Shahid S, Perera BJC (2018b) Cautionary note on the use of genetic programming in statistical downscaling. Int J Climatol 38:3449–3465 Salvi K, Ghosh S (2013) High-resolution multisite daily rainfall projections in India with statistical downscaling for climate change impacts assessment. J Geophys Res Atmos 118:3557–3578 Sehgal V, Lakhanpal A, Maheswaran R, Khosa R, Sridhar V (2018) Application of multi-scale wavelet entropy and multi-resolution Volterra models for climatic downscaling. J Hydrol 556:1078–1095 Sharma A (2000) Seasonal to interannual rainfall probabilistic forecasts for improved water supply management: part 1—a strategy for system predictor identification. J Hydrol 239:232–239 Spak S, Holloway T, Lynn B, Goldberg R (2007) A comparison of statistical and dynamical downscaling for surface temperature in North America. J Geophys Res Atmos 112:D08101. https://doi.org/10.1029/2005jd006712 Stanislawska K, Krawiec K, Kundzewicz ZW (2012) Modelling global temperature changes with genetic programming. Comput Math App 64:3717–3728 Stennett-Brown RK, Jones JJP, Stephenson TS, Taylor MA (2017) Future Caribbean temperature and rainfall extremes from statistical downscaling. Int J Climatol 37:4828–4845 Timbal B, Fernandez E, Li Z (2009) Generalization of a statistical downscaling model to provide local climate change projections for Australia. Environ Modell Softw 24:341–358 Vuillaume J, Hearth S (2018) Dynamic downscaling based on weather types classification: an application to extreme rainfall in South-East Japan. J Flood Risk Manag 11:e12340. https://doi.org/10.1111/jfr3.12340 Wang J, Swati FNU, Stein ML, Kotamarthi VR (2015) Model performance in spatiotemporal patterns of precipitation: new methods for identifying value added by a regional climate model. J Geophys Res Atmos 120:1239–1259 Wilby R, Dawson C, Barrow E (2002) SDSM—a decision support tool for the assessment of regional climate change impacts. Environ Modell Softw 17:145–157 Wu B, Wang J (2002) Winter arctic oscillation, Siberian High and East Asian winter monsoon. Geophys Res Lett 29:1897. https://doi.org/10.1029/2002gl015373 Yabusaki S, Tase N, Shimano Y (2010) Temporal variation of stable isotopes in precipitation at Tsukuba, Ogawa and Utsunomiya City in Japan. In: Taniguchi M, Holman IP (eds) Groundwater response to changing climate imprint. CRC Press, London, pp 55–66 Yang C, Wang N, Wang S (2017) A comparison of three predictor selection methods for statistical downscaling. Int J Climatol 37:1238–1249 Yato H, Nomura Y, Umehara K, Hosomichi A, Kawano S, Mouri H, Hagiya S (2017) Automated meteorological data acquisition system (AMeDAS) in Japan and field experiments to determine the effects of its observation environment. In: WMO international conference on automatic weather stations (ICAWS-2017). Offenbach am Main, Germany, 24–26 Oct 2017. https://www.wmo.int/pages/prog/www/IMOP/AWS-conference/ICAWS-2017_Topic_3.html. Accessed 12 April 2019