ABSTRACT
Machine learning models based on satellite data have been actively researched to serve as a proxy for the prediction of socio-economic development indicators. Such models have however rarely been tested for transferability over time, i.e. whether models learned on data for a certain year are able to make accurate predictions on data for another year. Using a dataset from the Indian census at two time points, for the years 2001 and 2011, we evaluate the temporal transferability of a simple machine learning model at sub-national scales of districts and propose a generic method to improve its performance. This method can be especially relevant when training datasets are small to train a robust prediction model. Then, we go further to build an aggregate development index at the district-level, on the lines of the Human Development Index (HDI) and demonstrate high accuracy in predicting the index based on satellite data for different years. This can be used to build applications to guide data-driven policy making at fine spatial and temporal scales, without the need to conduct frequent expensive censuses and surveys on the ground.
- 2019. Census of India Website: Office of the Registrar General and Census Commissioner, India. http://censusindia.gov.in/Google Scholar
- NITI Aayog. 2018. SDG India Index Baseline Report.Google Scholar
- Saikat Basu, Sangram Ganguly, Supratik Mukhopadhyay, Robert DiBiano, Manohar Karki, and Ramakrishna Nemani. 2015. Deepsat: a learning framework for satellite imagery. In Proceedings of the 23rd SIGSPATIAL international conference on advances in geographic information systems. ACM, 37.Google ScholarDigital Library
- Frank Bickenbach, Eckhardt Bode, Peter Nunnenkamp, and Mareike Söder. 2016. Night lights and regional GDP. Review of World Economics 152, 2 (2016), 425--447.Google ScholarCross Ref
- Gyanesh Chander, Brian L Markham, and Dennis L Helder. 2009. Summary of current radiometric calibration coefficients for Landsat MSS, TM, ETM+, and EO-1 ALI sensors. Remote sensing of environment 113, 5 (2009), 893--903.Google Scholar
- C Chandramouli and Registrar General. 2011. Census of India 2011. Provisional Population Totals. New Delhi: Government of India (2011).Google Scholar
- Nitesh V Chawla, Kevin W Bowyer, Lawrence O Hall, and W Philip Kegelmeyer. 2002. SMOTE: synthetic minority over-sampling technique. Journal of artificial intelligence research 16 (2002), 321--357.Google ScholarCross Ref
- Tianqi Chen and Carlos Guestrin. 2016. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. ACM, 785--794.Google ScholarDigital Library
- Xi Chen and William D Nordhaus. 2010. The value of luminosity data as a proxy for economic statistics. Technical Report. National Bureau of Economic Research.Google Scholar
- Zhaoxin Dai, Yunfeng Hu, and Guanhua Zhao. 2017. The suitability of different nighttime light data for GDP estimation at different spatial scales and regional levels. Sustainability 9, 2 (2017), 305.Google ScholarCross Ref
- Dave Donaldson and Adam Storeygard. 2016. The view from above: Applications of satellite data in economics. Journal of Economic Perspectives 30, 4 (2016), 171--98.Google ScholarCross Ref
- Eugenie Dugoua, Ryan Kennedy, and Johannes Urpelainen. 2018. Satellite data for the social sciences: measuring rural electrification with night-time lights. International journal of remote sensing 39, 9 (2018), 2690--2701.Google ScholarCross Ref
- Christopher D Elvidge, Kimberly E Baugh, Sharolyn J Anderson, Paul C Sutton, and Tilottama Ghosh. 2012. The Night Light Development Index (NLDI): a spatially explicit measure of human development from satellite data. Social Geography 7, 1 (2012), 23--35.Google ScholarCross Ref
- Swetava Ganguli, Jared Dunnmon, and Darren Hau. 2016. Predicting food security outcomes using cnns for satellite tasking.Google Scholar
- Bardan Ghimire, John Rogan, Víctor Rodríguez Galiano, Prajjwal Panday, and Neeti Neeti. 2012. An evaluation of bagging, boosting, and random forests for land-cover classification in Cape Cod, Massachusetts, USA. GIScience & Remote Sensing 49, 5 (2012), 623--643.Google ScholarCross Ref
- Giorgia Giovannetti, Elena Perra, et al. 2019. Syria in the Dark: Estimating the Economic Consequences of the Civil War through Satellite-Derived Night Time Lights. Technical Report. Universita'degli Studi di Firenze, Dipartimento di Scienze per l'Economia e âĂę.Google Scholar
- Ran Goldblatt, Alexis Rivera Ballesteros, and Jennifer Burney. 2017. High Spatial Resolution Visual Band Imagery Outperforms Medium Resolution Spectral Imagery for Ecosystem Assessment in the Semi-Arid Brazilian Sertão. Remote Sensing 9, 12 (2017), 1336.Google ScholarCross Ref
- Google. 2019. Landsat Algorithms. https://developers.google.com/earth-engine/landsatGoogle Scholar
- Dibyajyoti Goswami, Shyam Bihari Tripathi, Sansiddh Jain, Shivam Pathak, and Aaditeshwar Seth. 2019. Towards Building a District Development Model for India Using Census Data. (2019).Google Scholar
- Andrew Head, Mélanie Manguin, Nhat Tran, and Joshua E Blumenstock. 2017. Can Human Development be Measured with Satellite Imagery?. In ICTD. 8--1.Google Scholar
- Patrick Helber, Benjamin Bischke, Andreas Dengel, and Damian Borth. 2017. Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. arXiv preprint arXiv:1709.00029 (2017).Google Scholar
- J Vernon Henderson, Adam Storeygard, and David N Weil. 2012. Measuring economic growth from outer space. American economic review 102, 2 (2012), 994--1028.Google Scholar
- Tengyun Hu, Jun Yang, Xuecao Li, and Peng Gong. 2016. Mapping urban land use by using landsat images and open social data. Remote Sensing 8, 2 (2016), 151.Google ScholarCross Ref
- Wenjie Hu, Jay Harshadbhai Patel, Zoe-Alanah Robert, Paul Novosad, Samuel Asher, Zhongyi Tang, Marshall Burke, David Lobell, and Stefano Ermon. 2019. Mapping Missing Population in Rural India: A Deep Learning Approach with Satellite Imagery. arXiv preprint arXiv:1905.02196 (2019).Google Scholar
- Neal Jean, Marshall Burke, Michael Xie, W Matthew Davis, David B Lobell, and Stefano Ermon. 2016. Combining satellite imagery and machine learning to predict poverty. Science 353, 6301 (2016), 790--794.Google ScholarCross Ref
- David A Landgrebe. 2005. Signal theory methods in multispectral remote sensing. Vol. 29. John Wiley & Sons.Google Scholar
- Charlotta Mellander, José Lobo, Kevin Stolarick, and Zara Matheson. 2015. Night-time light data: A good proxy measure for economic activity? PloS one 10, 10 (2015), e0139779.Google ScholarCross Ref
- Brian Min and Kwawu Gaba. 2014. Tracking electrification in Vietnam using nighttime lights. Remote Sensing 6, 10 (2014), 9511--9529.Google ScholarCross Ref
- Brian Min, Kwawu Mensan Gaba, Ousmane Fall Sarr, and Alassane Agalassou. 2013. Detection of rural electrification in Africa using DMSP-OLS night lights imagery. International journal of remote sensing 34, 22 (2013), 8118--8141.Google Scholar
- KN Nischal, Radhika Radhakrishnan, Sanket Mehta, and Sumit Chandani. 2015. Correlating night-time satellite images with poverty and other census data of India and estimating future trends. In Proceedings of the Second ACM IKDD Conference on Data Sciences. ACM, 75--79.Google ScholarDigital Library
- Shailesh M Pandey, Tushar Agarwal, and Narayanan C Krishnan. 2018. Multi-task deep learning for predicting poverty from satellite images. In Thirty-Second AAAI Conference on Artificial Intelligence.Google ScholarCross Ref
- Anthony Perez, Swetava Ganguli, Stefano Ermon, George Azzari, Marshall Burke, and David Lobell. 2019. Semi-supervised multitask learning on multispectral satellite images using wasserstein generative adversarial networks (gans) for predicting poverty. arXiv preprint arXiv:1902.11110 (2019).Google Scholar
- Anthony Perez, Christopher Yeh, George Azzari, Marshall Burke, David Lobell, and Stefano Ermon. 2017. Poverty prediction with public landsat 7 satellite imagery and machine learning. arXiv preprint arXiv:1711.03654 (2017).Google Scholar
- Anupam Prakash, Avdhesh Kumar Shukla, Chaitali Bhowmick, and Robert Carl Michael Beyer. 2019. Night-time Luminosity: Does it Brighten Understanding of Economic Activity in India? (2019).Google Scholar
- United Nations Development Programme. 2019. Human Development Reports. http://hdr.undp.org/en/content/human-development-index-hdiGoogle Scholar
- Caleb Robinson, Fred Hohman, and Bistra Dilkina. 2017. A deep learning approach for population estimation from satellite imagery. In Proceedings of the 1st ACM SIGSPATIAL Workshop on Geospatial Humanities. ACM, 47--54.Google ScholarDigital Library
- Hans Rosling. 2019. Factfulness. Flammarion.Google Scholar
- SP Subash, Rajeev Ranjan Kumar, and KS Aditya. 2018. Satellite data and machine learning tools for predicting poverty in rural India. Agricultural Economics Research Review 31, 347-2019-571 (2018), 231--240.Google ScholarCross Ref
- Potnuru Kishen Suraj, Ankesh Gupta, Makkunda Sharma, Sourabh Bikash Paul, and Subhashis Banerjee. 2017. On monitoring development using high resolution satellite images. arXiv preprint arXiv:1712.02282 (2017).Google Scholar
- Binh Tang, Ying Sun, Yanyan Liu, and David S Matteson. 2018. Dynamic Poverty Prediction with Vegetation Index. (2018).Google Scholar
- USGS. 2019. Landsat Missions. https://www.usgs.gov/land-resources/nli/landsat/landsat-7?qt-science_support_page_related_con=0#qt-science_support_page_related_conGoogle Scholar
- USGS/Google. 2019. USGS Landsat 7 Collection 1 Tier 1 and Real-Time data TOA Reflectance. https://developers.google.com/earth-engine/datasets/catalog/LANDSAT_LE07_C01_T1_RT_TOAGoogle Scholar
- Gary R Watmough, Peter M Atkinson, Arupjyoti Saikia, and Craig W Hutton. 2016. Understanding the evidence base for poverty--environment relationships using remotely sensed satellite data: an example from Assam, India. World Development 78 (2016), 188--203.Google ScholarCross Ref
- Michael Xie, Neal Jean, Marshall Burke, David Lobell, and Stefano Ermon. 2016. Transfer learning from deep features for remote sensing and poverty mapping. In Thirtieth AAAI Conference on Artificial Intelligence.Google ScholarDigital Library
Index Terms
- Temporal Prediction of Socio-economic Indicators Using Satellite Imagery
Recommendations
Describing Patterns of Socio-Economic Development at Fine Spatial and Temporal Resolutions
CoDS COMAD 2020: Proceedings of the 7th ACM IKDD CoDS and 25th COMADOur vision is to describe, at fine spatial and temporal scales, the nuances of socio-economic development taking place as measured through a mix of data sources including censuses and surveys, satellite data, agricultural commodity prices, and ...
Quantitative mapping of pasture biomass using satellite imagery
A knowledge of the amount of pasture biomass available in farm paddocks is crucial for improving utilization and productivity in the Australian grazing industry. A method to quantitatively map the biomass of annual pastures under grazing has been ...
An improved high spatial and temporal data fusion approach for combining Landsat and MODIS data to generate daily synthetic Landsat imagery
Two key weaknesses of STDFA including sensor difference and spatial variability were adjusted.Three wildly used spatial and temporal fusion methods were compared.The correlation coefficient r had a negative exponential relationship with ratio of land ...
Comments