Skip to content

sampsonola/PyScript

Repository files navigation

PyScript: Hamoye Premier Project

Topic: Predictive Modeling of Policy Impact on Refugee Numbers

Introduction

The scale and severity of humanitarian displacements are on the rise. According to the United Nations High Commissioner for Refugees (UNHCR), the number of forcibly displaced persons worldwide reached 80 million in 2019, marking an increase of over 36 million since 2009. Displaced persons now constitute a growing proportion of the global population. The COVID-19 pandemic has further intensified displacement pressures, with projections indicating that by the end of 2021, the number of persons of concern will exceed 95 million. In response to these massive dislocations, humanitarian agencies are tasked with monitoring the situation and coordinating a response. Additionally, national governments must comprehend these migration patterns for political, administrative, and welfare planning purposes.

References and Ideas

  • UNHCR - Global Trends Report: UNHCR Website (This is the website of the United Nations High Commissioner for Refugees, a valuable resource for global refugee statistics and trends.)
  • World Bank - Forced Migration and Development: World Bank Resource This World Bank resource explores the links between forced migration and development.
  • Migration Policy Institute - Policy Beat: Migration Policy Institute This website from the Migration Policy Institute offers policy analysis and research on various migration topics

Recently, increased access to data and advancements in computational methods have enabled the construction of machine learning models for predicting displacement. This research aims to contribute to this progress by developing robust models that will assist decision-makers in identifying the most significant policies affecting refugee displacement.

Problem Statement

Current methods for understanding the impact of policies on refugee movements are often limited or rely on subjective analysis. This research project aims to develop a more objective and data-driven approach by utilizing machine learning models to predict how changes in national and international policies might influence refugee flows.

Objectives

  1. To analyze historical data from the UN Department of Economic and Social Affairs to identify trends and patterns in global refugee statistics.
  2. To investigate the effects of national and international policies on refugee flows using the provided dataset.
  3. To apply machine learning and neural network techniques such as Support Vector Machines (SVM), Logistic Regression and Long Short Time Memory (LSTM) to predict how changes in policy might influence refugee movements.
  4. To simulate the outcomes of potential policy modifications on refugee flows using the developed machine learning models.
  5. To provide policymakers and humanitarian organizations with data-driven insights to craft more effective and responsive refugee policies based on the research findings.

Methodologies

Dataset: UN Economic and Social Affair

Prediction problem definition and modeller decisions: The subsequent step involves structuring the prediction problem itself. We delineate nine crucial decisions that need to be made, including determining the unit of analysis, defining the time horizon, selecting the target variable, identifying the feature variables, addressing missing data and data quality issues, deciding on the modelling approach, outlining the model selection process, determining how to assess performance, and planning the deployment of the resulting models.

The summary of this approach can be seen in the table below:

Expected Outcomes

  1. Identification of Key Trends: The research will uncover key trends and patterns in global refugee statistics, providing insights into the factors influencing refugee flows.
  2. Prediction of Policy Effects: By applying machine learning techniques, the research will predict how changes in national and international policies might impact refugee movements.
  3. Simulation of Policy Modifications: The study will simulate the outcomes of potential policy modifications, allowing for an assessment of their potential effects on refugee flows.
  4. Data-Driven Policy Insights: The research findings will equip policymakers and humanitarian organisations with data-driven insights, enabling them to craft more effective and responsive refugee policies.
  5. Improved Policy Planning: The research outcomes will facilitate more informed and strategic decision-making in developing and implementing refugee policies at both national and international levels.

Conclusion

This project not only seeks to construct robust models for identifying the policies influencing refugee displacement but also to leverage these models to provide policymakers and humanitarian organizations with crucial data-driven insights. By understanding the potential impact of policies on refugee movements, stakeholders can develop more effective and responsive refugee policies, ultimately improving outcomes for refugees worldwide.

About

Hamoye Premier Project Proposal Topic: Predictive Modeling of Policy Impact on Refugee Numbers

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 5