Machine Learning at Central Banks

Chakraborty, Chiranjit; Joseph, Andreas

doi:10.2139/ssrn.3031796

Cited by 78 publications

(69 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ensemble methods, such as random forests or bootstrapped model averages, avoid this problem by reducing this variance, i.e. they lead to a better bias-variance trade-off (see also Chakraborty and Joseph (2017)).…”

Section: Decision Treesmentioning

confidence: 99%

Credit Growth, the Yield Curve and Financial Crisis Prediction: Evidence from a Machine Learning Approach

et al. 2020

Self Cite

View full text Add to dashboard Cite

We develop early warning models for financial crisis prediction using machine learning techniques on macrofinancial data for 17 countries over 1870-2016. Machine learning models mostly outperform logistic regression in out-of-sample predictions and forecasting. We identify economic drivers of our machine learning models using a novel framework based on Shapley values, uncovering non-linear relationships between the predictors and crisis risk. Throughout, the most important predictors are credit growth and the slope of the yield curve, both domestically and globally. A flat or inverted yield curve is of most concern when nominal interest rates are low and credit growth is high.

show abstract

Section: Decision Treesmentioning

confidence: 99%

Credit Growth, the Yield Curve and Financial Crisis Prediction: Evidence from a Machine Learning Approach

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…The main disadvantage of ANNs, commonly known as the black box criticism, is related to results' opacity and limited interpretability (see Han & Kamber, 2006, Angelini et al, 2008, Witten et al, 2011, Chakraborty & Joseph, 2017. However, as the black box criticism comes from a desire to tie down empirical estimation with an underlying economic theory (McNelis, 2005), this disadvantage is not an issue in our case: our only goal is to test whether electronic payment instruments data and an ANN are able to nowcast economic activity with fair precision -no underlying economic theory is to be tested.…”

Section: Prediction Methodsmentioning

confidence: 99%

“…the simple unweighted average. This is useful because the joint output of the committee of networks will usually achieve higher performance than any single network used in isolation (Bishop, 1995, Hagan et al, 2014, and because it enables to obtain a density forecast for the target variable (Chakraborty & Joseph, 2017).…”

Section: A Committee Of Artificial Neural Networkmentioning

confidence: 99%

Nowcasting economic activity with electronic payments data: A predictive modeling approach

León¹,

Ortega²

2018

View full text Add to dashboard Cite

The opinions contained in this document are the sole responsibility of the authors and do not commit Banco de la República or its Board of Directors. AbstractEconomic activity nowcasting (i.e. making current-period estimates) is convenient because most traditional measures of economic activity come with substantial lags. We aim at nowcasting ISE, a short-term economic activity indicator in Colombia. Inputs are ISE's lags and a dataset of payments made with electronic transfers and cheques among individuals, firms, and the central government. Under a predictive modeling approach, we employ a nonlinear autoregressive exogenous neural network model. Results suggest that our choice of inputs and predictive method enable us to nowcast economic activity with fair accuracy. Also, we validate that electronic payments data significantly reduces the nowcast error of a benchmark non-linear autoregressive neural network model. Nowcasting economic activity from electronic payment instruments data not only contributes to agents' decision making and economic modeling, but also supports new research paths on how to use retail payments data for appending current models.

show abstract

“…These are situations where the precise prediction of outcomes is important to inform decisions. 3 Examples include the forecasting of economic developments (Garcia et al (2017)), modelling the soundness of financial institutions (Chakraborty and Joseph (2017)), consumer credit scoring (Fuster et al (2017)), policy targeting based on uncertain outcomes (Andini et al (2017)), the prediction of extreme weather events in the face of climate change (Racah et al (2016)), medical image analysis and diagnosis (Litjens et al (2017)) or aiding expert judgement (Kleinberg et al (2018)).…”

Section: Introductionmentioning

confidence: 99%

Machine Learning at Central Banks

Cited by 78 publications

References 66 publications

Credit Growth, the Yield Curve and Financial Crisis Prediction: Evidence from a Machine Learning Approach

Credit Growth, the Yield Curve and Financial Crisis Prediction: Evidence from a Machine Learning Approach

Nowcasting economic activity with electronic payments data: A predictive modeling approach

Shapley Regressions: A Framework for Statistical Inference on Machine Learning Models

Contact Info

Product

Resources

About