{"title":"Integration of geospatial techniques and machine learning in land parcel prediction","authors":"Nekkanti Haripavan , Subhashish Dey , Chimakurthi Harika Mani Chandana","doi":"10.1016/j.geogeo.2025.100371","DOIUrl":null,"url":null,"abstract":"<div><div>The integration of geospatial techniques and machine learning algorithms has revolutionized our ability to analyze and predict changes in land parcels. In this research work leverage the power of Google Earth Engine to observe and interpret historical data spanning the last 2014–2023 years, in order to make informed predictions about future land parcel transformations. Our research will highlight the key components of this plan including data acquisition, preprocessing, feature engineering, and the application of machine learning models. We will explore how Google Earth Engine provides a robust platform for accessing vast geospatial datasets and performing complex analyses. By harnessing the temporal and spectral information captured by Earth observation satellites, we aim to identify patterns and trends in land parcel changes. These insights are used to train and fine-tune our machine learning models, which will subsequently forecast future land parcel developments. The project underscores the practical significance of our research work, as it can be applied to more domains such as urban planning, agriculture, forestry, and environmental monitoring. Furthermore, it showcases the potential of technology to enhance our understanding of the dynamic nature of our environment, and the role that predictive analytics plays in informed decision-making. One significant benefit is the feature selection that may be customized thanks to machine learning and geospatial approaches. Researchers and practitioners can customize their models by choosing the most pertinent variables for each land parcel forecasts from a wide range of spatial features. This flexibility guarantees that models can concentrate on the spatial features that have the biggest influence on the desired outcomes, improving the forecasts' overall performance and interpretability.</div></div>","PeriodicalId":100582,"journal":{"name":"Geosystems and Geoenvironment","volume":"4 2","pages":"Article 100371"},"PeriodicalIF":0.0000,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Geosystems and Geoenvironment","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2772883825000214","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The integration of geospatial techniques and machine learning algorithms has revolutionized our ability to analyze and predict changes in land parcels. In this research work leverage the power of Google Earth Engine to observe and interpret historical data spanning the last 2014–2023 years, in order to make informed predictions about future land parcel transformations. Our research will highlight the key components of this plan including data acquisition, preprocessing, feature engineering, and the application of machine learning models. We will explore how Google Earth Engine provides a robust platform for accessing vast geospatial datasets and performing complex analyses. By harnessing the temporal and spectral information captured by Earth observation satellites, we aim to identify patterns and trends in land parcel changes. These insights are used to train and fine-tune our machine learning models, which will subsequently forecast future land parcel developments. The project underscores the practical significance of our research work, as it can be applied to more domains such as urban planning, agriculture, forestry, and environmental monitoring. Furthermore, it showcases the potential of technology to enhance our understanding of the dynamic nature of our environment, and the role that predictive analytics plays in informed decision-making. One significant benefit is the feature selection that may be customized thanks to machine learning and geospatial approaches. Researchers and practitioners can customize their models by choosing the most pertinent variables for each land parcel forecasts from a wide range of spatial features. This flexibility guarantees that models can concentrate on the spatial features that have the biggest influence on the desired outcomes, improving the forecasts' overall performance and interpretability.