Outsmart your Strategies with Artificial Intelligence
Background:
Financial companies have widely adopted big data tools to uncover hidden patterns or unknown correlations that can help them make better predictions and thus more-informed business decisions. However, if not supported by rigorous methodological analyses, data mining can be more than misleading. The testing of systematic trading rules is usually done through backtesting and prone to spurious outperformance as a result of the data-mining bias. Multiple rules tested concurrently over the same history and optional stopping rules, along with some others, are commonly known as p-hacking.
Workshop Objective
Artificial Intelligence turned into a buzzword, an empty phrase. Overhyped, while underdelivered. Despite failed expectation management, there is value to find in AI for investment management. Experience a state-of-the-art workshop that provides a coherent framework and a useful set of tools to apply artificial intelligence with two widely-used programming languages: R and Python.
Learning Outcomes
- how to critically evaluate systematic trading strategies with the help of Monte Carlo simulations and artificial trading rules
- how to use some of the most well-known methods in machine learning to detect potential anomalies in empirical data
- how to benefit from a mixed cognitive architecture
Content:
Programming Languages Used: Python and R
Part I:
- Evaluate systematic trading strategies with Monte Carlo simulations and artificial trading rules in order to minimize the number of false positive trading strategies.
Part II:
- Study some of the most well-known methods in machine learning (i.e. random forest and neural networks) and learn how to use these tools to detect potential anomalies in empirical data.
- Develop your own automated trading strategies and augment them with artificial intelligence, notably genetic algorithms, to build smarter strategies.
Part III:
- Introduce the concept of mixed cognitive architecture, which provides a framework on how human and artificial systems can work together, in order to show how one can yield intelligent behavior in a diversity of complex environments.
Dr. Gregory Gadzinski is Senior Consultant at Panthera Solutions and also a full-time professor of Finance and Economics at the International University of Monaco, teaching a wide range of courses in the DBA, MBA and MFIN programs. He was previously an Assistant Professor of Economics at the Chair for International Economics in Cologne, Germany. Dr. Gadzinski was also a full-time researcher at the Hedge Fund Research Institute in Monaco. His consultancy experience includes mandates at ALPSTAR Management, a multi-strategy hedge fund and at the European Central bank, DG Research, Frankfurt, Germany. Dr. Gadzinski has a PhD from the Université de la Méditerranée, France, a postgraduate degree in Mathematical Economics and Econometrics and a “MagistèreIngénieurEconomiste” from the University Aix-Marseille II. He has published several scientific articles in prestigious journals such as the Journal of Asset Management, the Journal of Hedge Funds and Derivatives, and the Journal of Investing.
