Machine learning model to predict mental health crises from electronic health records

Publication in Nature Medicine by Roger Garriga ’17 and Javier Mas ’17 (Data Science) et al

Article cover in Nature Medicine

The use of machine learning in healthcare is still in its infancy. In this paper, we describe the project we did to predict psychotic episodes together with Birmingham’s psychiatric hospital. We hope to see these sorts of applications of ML in healthcare become the new standard in the future. The technology is ready, so it’s just a matter of getting it done!

Paper abstract

The timely identification of patients who are at risk of a mental health crisis can lead to improved outcomes and to the mitigation of burdens and costs. However, the high prevalence of mental health problems means that the manual review of complex patient records to make proactive care decisions is not feasible in practice. Therefore, we developed a machine learning model that uses electronic health records to continuously monitor patients for risk of a mental health crisis over a period of 28 days. The model achieves an area under the receiver operating characteristic curve of 0.797 and an area under the precision-recall curve of 0.159, predicting crises with a sensitivity of 58% at a specificity of 85%. A follow-up 6-month prospective study evaluated our algorithm’s use in clinical practice and observed predictions to be clinically valuable in terms of either managing caseloads or mitigating the risk of crisis in 64% of cases. To our knowledge, this study is the first to continuously predict the risk of a wide range of mental health crises and to explore the added value of such predictions in clinical practice.

(You can also read about the project in more detail in this article from UPF)

Citation

Garriga, R., Mas, J., Abraha, S. et al. Machine learning model to predict mental health crises from electronic health records. Nat Med (2022). https://doi.org/10.1038/s41591-022-01811-5

Connect with BSE authors

Roger Garriga ’17 is a Research Data Scientist at Koa Health. He is an alum of the BSE Master’s in Data Science.

Javier Mas ’17 is Lead Data Scientist at Kannact. He is an alum of the BSE Master’s in Data Science.

Application of bagging in day-ahead electricity price forecasting and factor augmentation

Publication in Energy Economics by Kadir Özen ’21 (PhD Track) and Dilem Yıldırım

Illustration of an energy plant and city lights with market graph

This paper has been published in the November 2021 issue of the journal Energy Economics.

Abstract

The electricity price forecasting (EPF) is a challenging task not only because of the uncommon characteristics of electricity but also because of the existence of many potential predictors with changing predictive abilities over time. In such an environment, how to account for all available factors and extract as much information as possible is the key to the production of accurate forecasts. To address this long-standing issue in a way that balances complexity and forecasting accuracy while facilitating the traceability of the predictor selection procedure, we propose the method of Bootstrap Aggregation (bagging). To forecast day-ahead electricity prices in a multivariate context for six major power markets, we construct a large-scale pure price model and apply the bagging approach in comparison with the popular Least Absolute Shrinkage and Selection Operator (LASSO) estimation method. Our forecasting study reveals that bagging provides substantial forecast improvements on daily and hourly scales in almost all markets over the popular LASSO estimation method. The differentiation in the forecast performances of the two approaches appears to arise, inter alia, from their structural differences in the explanatory variables selection process. Moreover, to account for the intraday hourly dependencies of day-ahead electricity prices, all our models are augmented with latent factors, and a substantial improvement is observed only in the forecasts from models covering a relatively limited number of predictors.

Highlights

  • We forecast day-ahead electricity prices for major markets with a large-scale model.
  • The method of Bootstrap Aggregation (bagging) is applied to generate forecasts.
  • Bagging appears to be very competitive and promising compared to the popular LASSO.
  • Factor augmentation is proposed to capture intraday hourly dependencies of prices.
  • Augmentation improves forecasts only for models with limited number of predictors.

Connect with BSE authors

portrait

Kadir Özen ’21 is a student in the PhD Program at UPF and BSE. He is an alum of the BSE Master’s in Economics and Finance (PhD Track).

Marta Morazzoni and Andrea Sy win EEA Young Economist Award

Their paper, “Female entrepreneurship, financial frictions and capital misallocation in the US,” has also been published in the Journal of Monetary Economics.

A photo of the two BSE alumni who wrote the award-winning paper, Marta Morazzoni and Andrea Sy, posing together at UPF Ciudatella Campus where they are both PhD candidates.
EEA Young Economist Awardees Marta Morazzoni ’18 and Andrea Sy ’18

BSE alumni Marta Morazzoni and Andrea Sy (both Economics Class of 2018) received the 2021 Young Economist Award from the European Economic Association and Unicredit Foundation for their paper, “Female entrepreneurship, financial frictions and capital misallocation in the US.”

The paper has also just been published in the Journal of Monetary Economics. (It originally appeared as a Barcelona School of Economics Working Paper.)

“New empirical evidence, highly relevant policy implications”

The EEA Young Economics award committee consisted of Philipp Kircher, Giacomo Ponzetto and Antonella Trigari. They noted that “the paper addresses an extremely important topic, offers new empirical evidence from micro-level data cleverly identifying informative moments, and builds a state-of-the-art general equilibrium model to rationalize the evidence and to provide highly relevant policy implications.”

Read the award committee’s report on the EEA website

Paper abstract

We document and quantify the effect of a gender gap in credit access on both entrepreneurship and input misallocation in the US. Female entrepreneurs are found to be more likely to face a rejection on their loan applications and to have a higher average product of capital, a sign of gender-driven capital misallocation that decreases in female-led firms’ access to finance. These results are not driven by differences in observable individual or businesses characteristics. Calibrating a heterogeneous agents model of entrepreneurship to the US economy, we show that the observed gap in credit access explains the bulk of the gender differences in capital allocation across firms. Eliminating such credit imbalance is estimated to potentially increase output by 4%, and to reduce capital misallocation by 12%.

Key findings

  • In the US, female entrepreneurs receive less business funding compared to male entrepreneurs.
  • Female-owned firms operate with lower levels of assets, resulting in gender-driven capital misallocation.
  • Female-led businesses are nonetheless relatively more profitable and have better credit risk scores.
  • Removing the gender gap in business financing is estimated to potentially increase output by 4%.

Connect with BSE authors

portrait

Marta Morazzoni ’18 is a PhD candidate at Universitat Pompeu Fabra and BSE. She is an alum of the BSE Master’s in Economics.

portrait

Andrea Sy ’18 is a PhD candidate at Universitat Pompeu Fabra and BSE. She is an alum of the BSE Master’s in Economics.

Economic gains from global cooperation in fulfilling climate pledges

Publication in Energy Policy by Sneha Thube ’16 (Economics) et al

Co2, Carbon Dioxide, Carbon, Oxygen, The Atmosphere
Image by Gerd Altmann from Pixabay

My paper “Economic gains from global cooperation in fulfilling climate pledges” (with Ruth Delzeitab and Christian H.C.A. Henning) is now available online.

Paper Abstract

Mitigation of CO2 emissions is a global public good that imposes different regional economic costs. We assess the distributional effects of cooperative versus non-cooperative CO2 markets to fulfil the Nationally Determined Contributions (NDCs), considering different CO2 permit allocation rules in cooperative markets. We employ a global computable general equilibrium model based on the GTAP-9 database and the add-on GTAP-Power database. Our results show the resulting winners and losers under different policy scenarios with different permit allocation rules. We see that in 2030, we can obtain gains as high as $106 billion from global cooperation in CO2 markets. A cooperative CO2 permit market with equal per capita allowances results in considerable monetary transfers from high per capita emission regions to low per capita emission regions. In per capita terms, these transfers are comparable to the Official Development Assistance (ODA) transfers. We also disaggregate the mitigation costs into direct and indirect shares. For the energy-exporting regions, the largest cost component is unambiguously the indirect mitigation costs.

Conclusions

With regard to the initial NDCs, aggregate economic gains from jointly achieving the NDCs are $106bn (i.e. 60% of costs with unilateral action) in 2030. Mobilizing cooperation via Article 6 is important.

When the costs are disaggregated into direct (i.e. domestic mitigation) and indirect (i.e. due to changes in international markets) within the energy-exporters (e.g., Russia, Canada, Middle East and North Africa) the dominant cost share arises from indirect costs.

We also model a scenario using where regional allowances allocated in proportion to the regional population (aka Carbon Egalitarianism) within a global ETS. This approach addresses global equity issues, aligns incentives of all countries & eliminates free-riding problem.

Large financial transfers (~$114bn in 2030) are generated via the carbon markets are leads to welfare improvements in the developing regions. These transfers are comparable to the per capita ODA received by some countries esp. in Sub-Saharan Africa.

The approach based on per capita emission benchmarking has also been suggested by Dr. Raghuram Rajan

If global justice is considered as a global public good, which similar to GHG mitigation, is underprovided, then the principle of carbon egalitarianism could promisingly combine an additional aspect to welfare, giving an important message for policymakers.

Connect with the author

portrait

Sneha Thube ’16 is a researcher at the Kiel Institute for World Economy. She is an alum of the Barcelona GSE Master’s in Economics.

Two Macro alumni publish in the same volume of European Economic Review

Publications by Nicolò Maffei-Faccioli ’15 and Alessandro Ruggieri ’12

The September 2021 volume of the journal European Economic Review includes two publications by alumni of the BSE Macroeconomic Policy and Financial Markets Program:

Does immigration grow the pie? Asymmetric evidence from Germany

by Nicolò Maffei-Faccioli ’15 (with Eugenia Vella)

We provide empirical evidence suggesting that net migration shocks can have substantial demand effects, potentially acting like positive Keynesian supply shocks. Using monthly administrative data (2006–2019) for Germany in a structural VAR, we show that the shocks stimulate vacancies, wages, house prices, consumption, investment, net exports, and output. Unemployment falls for natives (dominant job-creation effect), driving a decline in total unemployment, while rising for foreigners (dominant job-competition effect). The geographic origin of migrants and the education level of residents matter crucially for the transmission. Overall, the evidence implies that the policy debate should focus on redistributive strategies between natives and foreigners.

(Featured on this blog as a working paper last year)


Twin Peaks: Covid-19 and the labor market

by Alessandro Ruggieri ’12 (with Jake Bradley and Adam Hal Spencer)

This paper develops a choice-theoretic equilibrium model of the labor market in the presence of a pandemic. It includes heterogeneity in productivity, age and the ability to work from home. Worker and firm behavior changes in the presence of the virus, which itself has equilibrium consequences for the infection rate. The model is calibrated to the UK and counterfactual lockdown measures are evaluated. We find a different response in both the evolution of the virus and the labor market with different lockdown policies. A laissez-faire approach results in lives lost and acts as negative shock to the economy. A lockdown policy, absent any other intervention, will reduce the lives lost but increase the economic burden. Consistent with recent evidence, we find that the economic costs from lockdown are most felt by those earning the least. Finally, we introduce a job retention scheme as implemented by the UK Government and find that it spreads the economic hardship more equitably.


Connect with BSE authors

Nicolò and Alessandro are both alumni of the Barcelona School of Economics Master’s Program in Macroeconomic Policy and Financial Markets. They both got their PhDs from the IDEA Program (UAB and BSE).

portrait

Nicolò Maffei-Faccioli ’15 is a Senior Economist at Norges Bank.

 
portrait

Alessandro Ruggieri ’12 is an Assistant Professor at the University of Nottingham.

The additional costs of living with a disability in the UK

Publication in the European Journal of Health Economics by Lukas Schuelke ’21 (ITFD)

A woman in a wheelchair around Camden street market
Photo: iStock.com/VictorHuang

Last year I worked on the article, “Estimating the additional costs of living with a disability in the United Kingdom between 2013 and 2016,” which was based on my undergraduate dissertation and which just got published in the European Journal of Health Economics.

My co-authors Luke Munford and Marcello Morciano are affiliated with the School of Health Sciences at the University of Manchester.

Abstract

In the United Kingdom, more than 20% of the population live with a disability. Past evidence shows that being disabled is associated with functional limitations that often cause social exclusion and poverty. Therefore, it is necessary to analyse the connection between disability and poverty. This paper examines whether households with disabled members face extra costs of living to attain the same standard of living as their peers without disabled members. The modelling framework is based on the standard of living approach which estimates the extra income required to close the gap between households with and without disabled members. We apply an ordered logit regression to data from the Family Resources Survey between 2013 and 2016 to analyse the relationship between standard of living, income, and disability, conditional on other explanatory variables. We find that households with disabled members face considerable extra costs that go beyond the transfer payment of the government. The average household with disabled members saw their weekly extra costs continually increase from £293 in 2013 to £326 in 2016 [2020 prices]. Therefore, the government needs to adjust welfare policies to address the problem of extra costs faced by households with disabled members.

Connect with BSE authors

portrait

Lukas Schuelke ’21 is a Planning Analyst at Amazon in London, UK. He is an alum of the BSE Master’s in International Trade, Finance, and Development.

New evidence of granular business cycles from German cities

Federica Daniele ’13 shares a paper accepted to Review of Economics and Statistics.

journal cover

My paper “The Micro-Origins of Business Cycles: Evidence from German Metropolitan Areas” joint with Heiko Stueber has been accepted to the Review of Economics and Statistics. Here is a summary of our work:

Cities compete to attract large firms. When Amazon announced in 2017 the opening of its second headquarters, 238 US cities signed up for it. Large firms bring jobs and can boost local productivity through spillovers. However, the downside is that they generate excessive local volatility.

We leverage quarterly data on size of all German establishments from 1990 to 2014 to show that a buildup of concentration of economic activity in the hands of few sizable firms is systematically associated with higher volatility in local labor markets in subsequent months.

The reason is granularity. When concentration is high, shocks to large firms do not average out with shocks to smaller ones and the evolution of local employment ends up mimicking the evolution of employment in the large firm. The economy experiences “granular” business cycles.

Our paper is the first to provide solid time-series support to granular business cycles as in Carvalho and Grassi. However, we show that large firms do not seem capable to trigger both booms and busts alike. Our evidence points in favor of granularity-driven recessions only.

Finally, we calibrate the parameters governing local firm dynamics to match the local employment law of motion, because we want to see what are the causes of the disproportionate presence of large firms in big cities. We find that it’s because of higher growth opportunities in big cities.

Bottom line: the volatility externality imposed by large firms encourages short-time work schemes as opposed to layoffs and may justify using size-dependent forms of public support for crisis management, but the benefit might have to be weighed against potential moral hazard.

Connect with the author

portrait

Federica Daniele ’13 is an economist at the Bank of Italy. She holds a PhD from UPF and Barcelona GSE and is an alum of the Barcelona GSE Master’s in Economics.

Tackling domestic violence using large-scale empirical analysis

New paper in Journal of Empirical Legal Studies co-authored by Ria Ivandić ’13 (Economics)

A woman holds a sign in front of her face that reads, "Love shouldn't hurt."
Photo by Anete Lusina from Pexels

In England, domestic violence accounts for one-third of all assaults involving injury. A crucial part of tackling this abuse is risk assessment – determining what level of danger someone may be in so that they can receive the appropriate help as quickly as possible. It also helps to set priorities for police resources in responding to domestic abuse calls in times when their resources are severely constrained. In this research, we asked how we can improve on existing risk assessment, a research question that arose from discussions with policy makers who questioned the lack of systematic evidence on this.

Currently, the risk assessment is done through a standardised list of questions – the so-called DASH form (Domestic Abuse, Stalking and Harassment and Honour- Based Violence) – which consists of 27 questions that are used to categorise a case as standard, medium or high risk. The resulting DASH risk scores have limited power in predicting which cases will result in violence in the future.  Following this research, we suggest that a two-part procedure would do better both in prioritising calls for service and in providing protective resources to victims with the greatest need. 

In our predictive models, we use individual-level records on domestic abuse calls, crimes, victim and perpetrator data from the Greater Manchester Police to construct the criminal and domestic abuse history variables of the victim and perpetrator. We combine this with DASH questionnaire data in order to forecast reported violent recidivism for victim-perpetrator pairs.  Our predictive models are random forests, which are a machine-learning method consisting of a large number of classification trees that individually classify each observation as a predicted failure or non-failure. Importantly, we take the different costs of misclassification into account.  Predicting no recidivism when it actually happens (a false negative) is far worse in terms of social costs than predicting recidivism when it does not happen (a false positive). While we set the cost of incurring a false negative versus a false positive at 10:1, this is a parameter that can be adjusted by stakeholders. 

We show that machine-learning methods are far more effective at assessing which victims of domestic violence are most at risk of further abuse than conventional risk assessments. The random forest model based on the criminal history variables together with the DASH responses significantly outperforms the models based on DASH alone. The negative prediction error – that is, the share of cases that would be predicted not to have violence yet violence occurs in the future – is low at 6.3% as compared with an officer’s DASH risk score alone where the negative prediction error is 11.5%.  We also examine how much each feature contributes to the model performance. There is no single feature that clearly outranks all others in importance, but it is the combination of a wide variety of predictors, each contributing their own ‘insight’, which makes the model so powerful.

Following this research, we have been in discussion with police forces across the United Kingdom and policy makers working on the Domestic Abuse Bill to think how our findings could be incorporated in the response to domestic abuse. We hope this research acts as a building block to increasing the use of administrative datasets and empirical analysis to improve domestic violence prevention.

This post is based on the following article:

Grogger, J., Gupta, S., Ivandic, R. and Kirchmaier, T. (2021), Comparing Conventional and Machine-Learning Approaches to Risk Assessment in Domestic Abuse Cases. Journal of Empirical Legal Studies, 18: 90-130. https://doi.org/10.1111/jels.12276 

Media coverage

Connect with the author

Ria Ivandić ’13 is a Researcher at LSE’s Centre for Economic Performance (CEP). She is an alum of the Barcelona GSE Master’s in Economics.

Eliciting preferences for truth-telling in a survey of politicians

Publication in PNAS by Katharina Janezic ’16 (Economics) and Aina Gallego (IBEI and IPEG)

logo

Honesty is one of the most valued traits in politicians. Yet, because lies often remain undiscovered, it is difficult to study if some politicians are more honest than others. This paper examines which individual characteristics are correlated with truth-telling in a controlled setting in a large sample of politicians. We designed and embedded a game that incentivizes lying with a non-monetary method in a survey answered by 816 Spanish mayors. Mayors were first asked how interested they were in obtaining a detailed report about the survey results, and at the end of the survey, they had to flip a coin to find out whether they would be sent the report. Because the probability of heads is known, we can estimate the proportion of mayors who lied to obtain the report.

We find that a large and statistically significant proportion of mayors lied. Mayors that are members of the two major political parties lied significantly more. We further find that women and men were equally likely to lie. Finally, we find a negative relationship between truth-telling and reelection in the next municipal elections, which suggests that dishonesty might help politicians survive in office.

Connect with the authors

Cross-border effects of regulatory spillovers: Evidence from Mexico

Forthcoming JIE publication by Jagdish Tripathy ’11 (Economics)

Economics alum Jagdish Tripathy ’11 has a paper forthcoming in the Journal of International Economics on “Cross-border effects of regulatory spillovers: Evidence from Mexico.”

Paper abstract

This paper studies the spillover of a macroprudential regulation in Spain to the Mexican financial system via Mexican subsidiaries of Spanish banks. The spillover caused a drop in the supply of household credit in Mexico. Municipalities with a higher exposure to Spanish subsidiaries experienced a larger contraction in household credit. These localized contractions caused a drop in macroeconomic activity in the local non-tradable sector. Estimates of the elasticity of loan demand by the non-tradable sector to changes in household credit supply range from 1.2–1.8. These results emphasize cross-border effects of regulations in the presence of global banks.

Key takeaways

Loan-loss provisions introduced in Spain in 2012 imposed a significant burden on the balanced sheet of Spanish banks. This regulation was unrelated to the Mexican financial system or the credit conditions of Mexican households. However, Mexican subsidiaries of two large Spanish banks, BBVA and Santander, reduced lending to Mexican households in response to the regulation (Fig. 1).

figure
Fig. 1. Growth in credit lending by Spanish and non-Spanish banks in Mexico.

Mexican municipalities with a higher exposure to Spanish banks (Fig. 2) experienced a larger contraction in lending to households. This drop in lending to households (i.e. a drop in credit supply) was associated with a reduction in lending to the local non-tradable sector driven by a drop in local demand. This shows (1) cross-border effects of a macroprudential regulation on lending and economic activity, and (2) the macroeconomic effects of shocks in lending to households in an emerging economy.

map
Fig. 2. Share of Spanish banks in the household credit market across Mexican municipalities.

About the author

Jagdish Tripathy ’11 is an Advisor at Bank of England. He is an alum of the Barcelona GSE Master’s in Economics and has his PhD from GPEFM (UPF and Barcelona GSE).

LinkedIn | Twitter | Website