Making useful conflict predictions
Methods for addressing skewed classes and implementing cost-sensitive learning in the study of state failure
Bibliographic Data
| ID | 3917145 |
|---|---|
| Authors | Ryan Kennedy (0000-0003-1881-6887, University of Houston, corresponding author) |
| Year | 2015 |
| Volume | 52 |
| Issue | 5 |
| Pages | 649-664 |
| Publication date | 2015-09-01 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Journal of Peace Research (JOURNAL) |
| Journal identifiers | ISSN: 0022-3433 • E-ISSN: 1460-3578 |
| Publisher | SAGE Publications Inc (PUBLISHER) |
| DOI | 10.1177/0022343315585611 |
| OpenAlex | W2193174802 |
| Language | EN |
| Citations received | 3 |
| References cited | 40 |
One of the major issues in predicting state failure is the relatively rare occurrence of event onset. This class skew problem can cause difficulties in both estimating a model and selecting a decision boundary. Since the publication of King & Zeng's studies in 2001, scholars have utilized case-control methods to address this issue. This article builds on the landmark research of the Political Instability Task Force comparing the case-control approach to several other methods from the machine learning field and some original to this study. Case-control methods have several practical disadvantages and show no measurable advantages in prediction. The article also introduces cost-sensitive methods for determining a decision boundary. This explication raises questions about the Task Force's formulation of a decision boundary and suggests methods for making useful predictions for policy. I find that the decision boundary chosen by the PITF implicitly assumes that the cost of intervention is about 7.7% of the cost of non-intervention when state failure will take place. These findings demonstrate that there is still much work to be done in predicting state failure, especially in limiting the number of false positives. More generally, it suggests caution in using accuracy as a measure of success when significant class imbalance exists in the data
Boundary (topology) · Class (philosophy) · Data mining · Decision boundary · Event (particle physics) · Explication · False positive paradox · Machine learning · Operations research · Risk analysis (engineering) · Skew · Task (project management) · Artificial Intelligence · Computer Science · Corruption and Economic Development · Engineering · Mathematics · Political Conflict and Governance · Qualitative Comparative Analysis Research
Semiparametric Regression for the Social Sciences
Time Series Analysis for the Social Sciences
Data Analysis Using Regression and Multilevel/Hierarchical Models
Bayesian Analysis for the Social Sciences
An introduction to ROC analysis
Comparing the Areas under Two or More Correlated Receiver Operating Characteristic Curves
Neonatal, postneonatal, childhood, and under-5 mortality for 187 countries, 1970–2010
Smote
Breaking the Conflict Trap: Civil War and Development Policy
Greed and grievance in civil war
Resort to War, 1816-2007
Crisis Early Warning and Decision Support
Improving Forecasts of State Failure
Modeling Heterogeneous Treatment Effects in Survey Experiments with Bayesian Additive Regression Trees
Which Elections Can Be Lost
Logistic Regression in Rare Events Data
Competing Approaches to Predicting Supreme Court Decision Making
Predicting Armed Conflict, 2010-20501
Toward the Development of an Integrated Crisis Warning System
Explaining Rare Events in International Relations
A Political Economy of Aid
Pitfalls and Prospects in the Peacekeeping Literature
Forecasting Forpolicymaking in Thepost–coldwarperiod
A Global Model for Forecasting Political Instability
State-Level Forecasts of the 2012 Federal and Gubernatorial Elections
Analyzing Incomplete Political Science Data
Tantalus on the Road to Asymptopia
Factors relevant to the validity of experiments in social settings
Forecasting the onset of genocide and politicide
Forecasting is difficult, especially about the future
The perils of policy by p-value
Seven deadly sins of contemporary quantitative political analysis
Improving Quantitative Studies of International Conflict
Untangling Neural Nets
| Unique citing works | 3 |
|---|---|
| Citations per year | 0,33 |
| Citation span | 2017 - 2018 (2) |
| Citation velocity | historical |
| Highly cited | No |
| Citation types | Neutral: 2 |