Periodic Reporting for period 1 - PaCE (The Patterns of Conflict Emergence: Developing an Automated Pattern Recognition System for Conflict)
Okres sprawozdawczy: 2022-01-01 do 2023-06-30
The PaCE Project (Patterns of Conflict Emergence).
There have been more than 200 wars since the start of the 20th century, leading to 35 million battle deaths and countless more civilian casualties. Large-scale political violence still kills hundreds every day across the world. International conflicts and civil wars also lead to forced migration, disastrous economic consequences, weakened political systems, and poverty.
The recurrence of wars despite their tremendous economic, social, and institutional costs, may suggest that we are doomed to repeat the errors of the past. Does history indeed repeat itself? Are there particular temporal patterns which occur in the build up to the onset of war? Would better understanding of these patterns help us to avoid such conflicts? This project aims to uncover, cluster, and classify such patterns in meaningful ways to help us improve future forecasts. In particular, it addresses three related central research questions:
1. What are the recurring pre-conflict patterns? Certain indicators may follow a typical path—a motif— prior to conflict events (inter- or intra-state). Or are the variables associated with conflict chaotic and therefore inherently unpredictable? Using novel methods in social sciences, we search for patterns in the observable actions that international leaders and actors take prior to conflict events, as well as in their perceptions. This will be done at multiple levels of resolution—the minute, the month, the year—and using original data on financial assets, news articles, and diplomatic cables.
2. Can we exploit these patterns for prediction purposes? We will cluster and classify sequences to understand where tensions are headed—escalation, diffusion, or decline? Answers to these questions could improve our ability to forecast wars by building an automated pattern recognition system for conflict and geopolitical crises. It will also help political scientists to answer theoretical questions that have so far remained out of reach because of the limits of existing methodological approaches and data.
3. Finally, we will apply the patterns to the generation of new theories about conflict processes by identifying the key sequences and combinations thereof that are particularly dangerous. Most work involving complex dynamics remains mostly theoretical, mostly because complex, non-linear dynamics involving escalation or diffusion are difficult to study empirically with existing methods.
This project aims to answer three main questions: (i) Are there patterns at all in conflict escalation? Are time-series associated with conflict purely chaotic and complex, or do they exhibit unexploited redundancy that can be used for predictions? (b) What do those patterns look like, and can their shapes inform our theoretical understanding of conflict and escalation? (c) Can we exploit these shapes to classify sequences of events as conflictual or peaceful for the purpose of improving our forecasts?
To our knowledge, no current system directly addresses these questions. Real-time crisis monitoring systems have emerged both in academia and in the policy area, but they rely on methods which (i) do not attempt to measure how fundamentally chaotic (and therefore predictable) their data is, and (ii) do not treat entire sequences as units of analysis and hence may fail to identify important geometric shapes and redundancies in the data.
(Grant agreement no:101002240)
There have been more than 200 wars since the start of the 20th century, leading to 35 million battle deaths and countless more civilian casualties. Large-scale political violence still kills hundreds every day across the world. International conflicts and civil wars also lead to forced migration, disastrous economic consequences, weakened political systems, and poverty.
The recurrence of wars despite their tremendous economic, social, and institutional costs, may suggest that we are doomed to repeat the errors of the past. Does history indeed repeat itself? Are there particular temporal patterns which occur in the build up to the onset of war? Would better understanding of these patterns help us to avoid such conflicts? This project aims to uncover, cluster, and classify such patterns in meaningful ways to help us improve future forecasts. In particular, it addresses three related central research questions:
1. What are the recurring pre-conflict patterns? Certain indicators may follow a typical path—a motif— prior to conflict events (inter- or intra-state). Or are the variables associated with conflict chaotic and therefore inherently unpredictable? Using novel methods in social sciences, we search for patterns in the observable actions that international leaders and actors take prior to conflict events, as well as in their perceptions. This will be done at multiple levels of resolution—the minute, the month, the year—and using original data on financial assets, news articles, and diplomatic cables.
2. Can we exploit these patterns for prediction purposes? We will cluster and classify sequences to understand where tensions are headed—escalation, diffusion, or decline? Answers to these questions could improve our ability to forecast wars by building an automated pattern recognition system for conflict and geopolitical crises. It will also help political scientists to answer theoretical questions that have so far remained out of reach because of the limits of existing methodological approaches and data.
3. Finally, we will apply the patterns to the generation of new theories about conflict processes by identifying the key sequences and combinations thereof that are particularly dangerous. Most work involving complex dynamics remains mostly theoretical, mostly because complex, non-linear dynamics involving escalation or diffusion are difficult to study empirically with existing methods.
This project aims to answer three main questions: (i) Are there patterns at all in conflict escalation? Are time-series associated with conflict purely chaotic and complex, or do they exhibit unexploited redundancy that can be used for predictions? (b) What do those patterns look like, and can their shapes inform our theoretical understanding of conflict and escalation? (c) Can we exploit these shapes to classify sequences of events as conflictual or peaceful for the purpose of improving our forecasts?
To our knowledge, no current system directly addresses these questions. Real-time crisis monitoring systems have emerged both in academia and in the policy area, but they rely on methods which (i) do not attempt to measure how fundamentally chaotic (and therefore predictable) their data is, and (ii) do not treat entire sequences as units of analysis and hence may fail to identify important geometric shapes and redundancies in the data.
(Grant agreement no:101002240)
1.1 CASUAL INFERENCE
1.1.1 Accounting for different speeds in comparative case studies: Dynamic Synthetic Controls
Synthetic controls are widely used to estimate the causal effect of a treatment. However, they do not account for the different speeds at which units respond to changes. Reactions may be inelastic or `sticky’ and thus slower due to varying regulatory, institutional, or political environments. We show that these discrepancies in reaction speeds can lead to biased estimates of causal effects. To address this issue, we introduce a dynamic synthetic control approach that accommodates varying speeds in time series, resulting in improved synthetic control estimates. We apply our method to re-estimate the effects of terrorism on income (Abadie and Gardeazabal, 2003), tobacco laws on consumption (Abadie, Diamond, and Hainmueller 2010), and German reunification on GDP (Abadie, Diamond, and Hainmueller 2015. We find that our approach reduces errors in the estimates of true treatment effects by up to 70% compared to traditional synthetic controls, improving our ability to make robust inferences.
1.1.2 Accounting for different speeds in regression analysis
This project aims to minimise potential bias arising from speed differences. By correcting for varying speeds across variables, causal effects are estimate with more accuracy.
1.1.3 Matching with varying histories (TBC)
1.2 Patterns in Time Series
1.2.1 Leveraging Temporal Patterns in Forecasting
Recurring temporal patterns naturally emerge from underlying processes and interactions in a variety of disciplines, ranging from epidemiology and ecology to social sciences and physics. These patterns and motifs hold considerable promise for enhancing the precision of time-series forecasting. This study introduces an innovative method that not only identifies these repeating patterns but also incorporates them as dynamic covariates in traditional time-series forecasting models. By leveraging time series clustering techniques our approach actively seeks out recurring patterns in time series data, transforming them into dynamic covariates that augment prediction capabilities. Our methodology is evaluated with three widely used forecasting models (ARIMA, RF and LSTM). Each model is implemented in its standard form and subsequently augmented with dynamic covariates. The results convincingly demonstrate that the introduction of dynamic covariates significantly improves the prediction accuracy across all three models. Our findings underline the potential of adding recurring motifs to prediction tasks for a variety of algorithms.
2. TEMPORAL PATTERNS IN SOCIAL PHENOMENA
2.1 Temporal Patterns in Protest Events (TBC)
2.2 Temporal Patterns in Migration Flows
Existing work has contributed to a rich understanding of the factors that affect why and when people leave, at the micro, the meso-, and the macro-level. What is less understood are the dynamics of migration flows over time. Existing work typically focuses on static variables at the country-year level, and the temporal dynamics are rarely investigated. Are there recurring temporal patterns in migration flows? And can we use these patterns to improve our forecasts of the number of migrants? Here, we develop new methods to uncover motifs in the number of migrants over time, and use these motifs for forecasting. By clustering the time series into common shapes that allow for different time scales, we show that the inclusion of temporal clusters does improve our ability to forecast migrant flows. We apply the new method to the case of South Sudan.
3. FORECASTING
(TBC)
4. DATA
4.1 Monitoring Conflict from Space
(TBC)
1.1.1 Accounting for different speeds in comparative case studies: Dynamic Synthetic Controls
Synthetic controls are widely used to estimate the causal effect of a treatment. However, they do not account for the different speeds at which units respond to changes. Reactions may be inelastic or `sticky’ and thus slower due to varying regulatory, institutional, or political environments. We show that these discrepancies in reaction speeds can lead to biased estimates of causal effects. To address this issue, we introduce a dynamic synthetic control approach that accommodates varying speeds in time series, resulting in improved synthetic control estimates. We apply our method to re-estimate the effects of terrorism on income (Abadie and Gardeazabal, 2003), tobacco laws on consumption (Abadie, Diamond, and Hainmueller 2010), and German reunification on GDP (Abadie, Diamond, and Hainmueller 2015. We find that our approach reduces errors in the estimates of true treatment effects by up to 70% compared to traditional synthetic controls, improving our ability to make robust inferences.
1.1.2 Accounting for different speeds in regression analysis
This project aims to minimise potential bias arising from speed differences. By correcting for varying speeds across variables, causal effects are estimate with more accuracy.
1.1.3 Matching with varying histories (TBC)
1.2 Patterns in Time Series
1.2.1 Leveraging Temporal Patterns in Forecasting
Recurring temporal patterns naturally emerge from underlying processes and interactions in a variety of disciplines, ranging from epidemiology and ecology to social sciences and physics. These patterns and motifs hold considerable promise for enhancing the precision of time-series forecasting. This study introduces an innovative method that not only identifies these repeating patterns but also incorporates them as dynamic covariates in traditional time-series forecasting models. By leveraging time series clustering techniques our approach actively seeks out recurring patterns in time series data, transforming them into dynamic covariates that augment prediction capabilities. Our methodology is evaluated with three widely used forecasting models (ARIMA, RF and LSTM). Each model is implemented in its standard form and subsequently augmented with dynamic covariates. The results convincingly demonstrate that the introduction of dynamic covariates significantly improves the prediction accuracy across all three models. Our findings underline the potential of adding recurring motifs to prediction tasks for a variety of algorithms.
2. TEMPORAL PATTERNS IN SOCIAL PHENOMENA
2.1 Temporal Patterns in Protest Events (TBC)
2.2 Temporal Patterns in Migration Flows
Existing work has contributed to a rich understanding of the factors that affect why and when people leave, at the micro, the meso-, and the macro-level. What is less understood are the dynamics of migration flows over time. Existing work typically focuses on static variables at the country-year level, and the temporal dynamics are rarely investigated. Are there recurring temporal patterns in migration flows? And can we use these patterns to improve our forecasts of the number of migrants? Here, we develop new methods to uncover motifs in the number of migrants over time, and use these motifs for forecasting. By clustering the time series into common shapes that allow for different time scales, we show that the inclusion of temporal clusters does improve our ability to forecast migrant flows. We apply the new method to the case of South Sudan.
3. FORECASTING
(TBC)
4. DATA
4.1 Monitoring Conflict from Space
(TBC)
Overall, PaCE will provide researchers with a) new fine-grained data; b) a set of tools to better understand conflicts; c) extracted features to build better theories; d) improved forecasts; and e) answers to fundamental questions about whether history repeats itself, and the inherent predictability of conflict processes.