Big Data and Probably Approximately Correct Learning In The Presence of Noise: Implications For Financial Risk Management

© 2019 [International Journal of Artificial Intelligence]. High accuracy forecasts are essential to financial risk management, where machine learning algorithms are frequently employed. We derive a new theoretical bound on the sample complexity for Probably Approximately Correct (PAC) learning in the presence of noise, and does not require specification of the hypothesis set |H|. We demonstrate that for realistic financial applications where |H| is typically infinite. This is contrary to prior theoretical conclusions. We further show that noise, which is a non-trivial component of big data, has a dominating impact on the data size required for PAC learning. Consequently, contrary to current big data trends, we argue that high quality data is more important than large volumes of data. This paper additionally demonstrates that the level of algorithmic sophistication, specifically the Vapnik-Chervonenkis (VC) dimension, needs to be traded-off against data requirements to ensure optimal algorithmic performance. Finally, our new Theorem can be applied to a wider range of machine learning algorithms, as it does not impose finite |H| requirements. This paper contributes to theoretical and applied research in the domain of machine learning for financial applications.

[1]  Hans Ulrich Simon,et al.  An Almost Optimal PAC Algorithm , 2015, COLT.

[2]  Stephen P. Ryan,et al.  Machine Learning Methods for Demand Estimation , 2015 .

[3]  David Bholat,et al.  Big Data and central banks , 2015, Big Data & Society.

[4]  Xingquan Zhu,et al.  Class Noise vs. Attribute Noise: A Quantitative Study , 2003, Artificial Intelligence Review.

[5]  Hal R. Varian,et al.  Big Data: New Tricks for Econometrics , 2014 .

[6]  Piet Daas,et al.  Big Data as a Source for Official Statistics , 2015 .

[7]  A. Engel,et al.  VAPNIK-CHERVONENKIS DIMENSION OF NEURAL NETWORKS WITH BINARY WEIGHTS , 1996, cond-mat/9608156.

[8]  Nikos D. Lagaros,et al.  AIRMS: A risk management tool using machine learning , 2018, Expert Syst. Appl..

[9]  Peter L. Bartlett,et al.  Neural Network Learning - Theoretical Foundations , 1999 .

[10]  Eyal Kushilevitz,et al.  PAC learning with nasty noise , 1999, Theoretical Computer Science.

[11]  Francisco Javier Población García.  Exchange Rate Risk , 2017 .

[12]  Dana Angluin,et al.  Learning from noisy examples , 1988, Machine Learning.

[13]  Pradeep Sen,et al.  Implementation of Random Parameter Filtering , 2011 .

[14]  Steven Finlay,et al.  Predictive Analytics, Data Mining and Big Data: Myths, Misconceptions and Methods , 2014 .

[15]  Vladimir Vapnik,et al.  Chervonenkis: On the uniform convergence of relative frequencies of events to their probabilities , 1971 .

[16]  Thomas F. Coleman,et al.  Moment matching machine learning methods for risk management of large variable annuity portfolios , 2018 .

[17]  Michael Kearns,et al.  Efficient noise-tolerant learning from statistical queries , 1993, STOC.

[18]  Pamela K. Coats,et al.  A neural network for classifying the financial health of a firm , 1995 .

[19]  Sendhil Mullainathan,et al.  Machine Learning: An Applied Econometric Approach , 2017, Journal of Economic Perspectives.

[20]  Murtaza Haider,et al.  Beyond the hype: Big data concepts, methods, and analytics , 2015, Int. J. Inf. Manag..

[21]  Victor Chernozhukov,et al.  Post-Selection Inference for Generalized Linear Models With Many Controls , 2013, 1304.3969.

[22]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[23]  Rafael Pino-Mejías,et al.  Credit scoring models for the microfinance industry using neural networks: Evidence from Peru , 2013, Expert Syst. Appl..

[24]  Hans-Peter Kriegel,et al.  A survey on unsupervised outlier detection in high‐dimensional numerical data , 2012, Stat. Anal. Data Min..

[25]  A. Kyle,et al.  The Flash Crash: High-Frequency Trading in an Electronic Market , 2017 .

[26]  Leslie G. Valiant,et al.  A theory of the learnable , 1984, CACM.