skip to main content

Explore Research Products in the NSF-PAR

Title: Applying machine learning to social datasets: a study of migration in southwestern Bangladesh using random forests
Abstract

As researchers collect large amounts of data in the social sciences through household surveys, challenges may arise in how best to analyze such datasets, especially where motivating theories are unclear or conflicting. New analytical methods may be necessary to extract information from these datasets. Machine learning techniques are promising methods for identifying patterns in large datasets, but have not yet been widely used to identify important variables in social surveys with many questions. To demonstrate the potential of machine learning to analyze large social datasets, we apply machine learning techniques to the study of migration in Bangladesh. The complexity of migration decisions makes them suitable for analysis with machine learning techniques, which enable pattern identification in large datasets with many covariates. In this paper, we apply random forest methods to analyzing a large survey which captures approximately 2000 variables from approximately 1700 households in southwestern Bangladesh. Our analysis ranked the covariates in the dataset in terms of their predictive power for migration decisions. The results identified the most important covariates, but there exists a tradeoff between predictive ability and interpretability. To address this tradeoff, random forests and other machine learning algorithms may be especially useful in combination with more traditional regression methods. To develop insights into how the important variables identified by the random forest algorithm impact migration, we performed a survival analysis of household time to first migration. With this combined analysis, we found that variables related to wealth and household composition are important predictors of migration. Such multi-methods approaches may help to shed light on factors contributing to migration and non-migration.

  more » « less
Award ID(s):
1716909
NSF-PAR ID:
10368154
Author(s) / Creator(s):
; ; ; ; ;
Publisher / Repository:
Springer Science + Business Media
Date Published:
Journal Name:
Regional Environmental Change
Volume:
22
Issue:
2
ISSN:
1436-3798
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ( , Ambio)
    null (Ed.)
    Abstract Non-migration is an adaptive strategy that has received little attention in environmental migration studies. We explore the leveraging factors of non-migration decisions of communities at risk in coastal Bangladesh, where exposure to both rapid- and slow-onset natural disasters is high. We apply the Protection Motivation Theory (PMT) to empirical data and assess how threat perception and coping appraisal influences migration decisions in farming communities suffering from salinization of cropland. This study consists of data collected through quantitative household surveys ( n  = 200) and semi-structured interviews from four villages in southwest coastal Bangladesh. Results indicate that most respondents are unwilling to migrate, despite better economic conditions and reduced environmental risk in other locations. Land ownership, social connectedness, and household economic strength are the strongest predictors of non-migration decisions. This study is the first to use the PMT to understand migration-related behaviour and the findings are relevant for policy planning in vulnerable regions where exposure to climate-related risks is high but populations are choosing to remain in place. 
    more » « less
  2. ; ( , ICPSR - Interuniversity Consortium for Political and Social Research)

    The Bangladesh Environment and Migration Survey (BEMS) collects detailed retrospective information about migration trips in southwest Bangladesh, including the first, last, and second-to-last to internal destinations, India, and other international destinations. BEMS collects information about the year, origin, destination, and duration of all trips. Furthermore, BEMS includes information on migration and livelihood histories, socioeconomic conditions, agricultural resources and practices, disasters and perceptions about environment, and self-reported health.

    Dataset 1 is a household-level file with information about household composition, economic and migratory activity of household members, land ownership/usage, business ownership, household environmental perceptions, environmental conditions, agricultural activities, and physical and psychological health/well-being of household members. Dataset 2 is an individual-level file containing details of internal and international migration trips, as well as measures of economic and social activity during those trips. It also contains information provided by household heads, spouses, and other migrants in the household. Dataset 3 is an individual-level data file that provides general demographic information and brief migration history for each member of a surveyed household. It also includes health information for the head of household and spouse.

    The purpose of the Bangladesh Environment and Migration Survey (BEMS) is to understand patterns and processes of contemporary internal and international migration in Bangladesh. The project derives from a multi-disciplinary research effort that will generate data on the characteristics and behavior of Bangladeshi migrants and non-migrants and the communities in which they live, and examine whether and how environmental stressors (e.g., salinity, riverbank erosion) affect patterns of migration in this region. The household ethnosurvey is administered to self-identified household heads and spouses in randomly selected households. After gathering social, demographic, and economic information on households and their members, interviewers will collect basic information on each person's first, 2nd to last, and last (or most recent) internal and international migration trips. From household heads and spouses, they will compile migration histories and administer a detailed series of questions about a selection of these trips, focusing on economic livelihoods, methods of moving, connections to other migrants, and use of health and school services. In addition to detailed migration histories, the BEMS will collect information about household wealth, physical conditions of households and communities, and perceptions of environmental conditions. It will also gather some self-reported health information about household members, such as recent illnesses, use of health services, height and weight, and diet. The BEMS is closely modeled on the sampling design and ethnosurvey used in the Mexican Migration Project. The BEMS data were collected in 20 research sites from a random sample of 200 households in each site in 2019. BEMS data include a total of 4,000 households in communities broadly covering the southwest region of Bangladesh. Households in southwest Bangladesh. Smallest Geographic Unit: Administrative region

    For more information about this study, please visit the ISEE Bangladesh project website.

     
    more » « less
  3. ; ; ; ( , Frontiers in Genetics)
    Random forest is considered as one of the most successful machine learning algorithms, which has been widely used to construct microbiome-based predictive models. However, its use as a statistical testing method has not been explored. In this study, we propose “Random Forest Test” (RFtest), a global (community-level) test based on random forest for high-dimensional and phylogenetically structured microbiome data. RFtest is a permutation test using the generalization error of random forest as the test statistic. Our simulations demonstrate that RFtest has controlled type I error rates, that its power is superior to competing methods for phylogenetically clustered signals, and that it is robust to outliers and adaptive to interaction effects and non-linear associations. Finally, we apply RFtest to two real microbiome datasets to ascertain whether microbial communities are associated or not with the outcome variables. 
    more » « less
  4. ; ; ( , Digital Discovery)
    Obtaining useful insights from machine learning models trained on experimental datasets collected across different groups to improve the sustainability of chemical processes can be challenging due to the small size and heterogeneity of the dataset. Here we show that shallow learning models such as decision trees and random forest algorithms can be an effective tool for guiding experimental research in the sustainable chemistry field. This study trained four different machine learning algorithms (linear regression, decision tree, random forest, and multilayer perceptron) using different sized datasets containing up to 520 unique reaction conditions for the nitrogen reduction reaction (NRR) on heterogeneous electrocatalysts. Using the catalyst properties and experimental conditions as the features, we determined the ability of each model to regress the ammonia production rate and the faradaic efficiency. We observed that the shallow learning decision tree and random forest models had equal or better predictive power compared to the deep learning multilayer perceptron models and the simple linear regression models. Moreover, decision tree and random forest models enable the extraction of feature importance, which is a powerful tool in guiding experimental research. Analysis of the models showed the complex interaction between the applied potential and catalysts on the effective rate for the NRR. We also suggest some underexplored catalysts–electrolyte combinations to experimental researchers looking to improve both the rate and efficiency of the NRR reaction. 
    more » « less
  5. ; ; ; ; ; ; ( , Molecular Ecology)
    Abstract

    Predictive phylogeography seeks to aggregate genetic, environmental and taxonomic data from multiple species in order to make predictions about unsampled taxa using machine‐learning techniques such as Random Forests. To date, organismal trait data have infrequently been incorporated into predictive frameworks due to difficulties inherent to the scoring of trait data across a taxonomically broad set of taxa. We refine predictive frameworks from two North American systems, the inland temperate rainforests of the Pacific Northwest and the Southwestern Arid Lands (SWAL), by incorporating a number of organismal trait variables. Our results indicate that incorporating life history traits as predictor variables improves the performance of the supervised machine‐learning approach to predictive phylogeography, especially for the SWAL system, in which predictions made from only taxonomic and climate variables meets only moderate success. In particular, traits related to reproduction (e.g., reproductive mode; clutch size) and trophic level appear to be particularly informative to the predictive framework. Predictive frameworks offer an important mechanism for integration of organismal trait, environmental data, and genetic data in phylogeographic studies.

     
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email