Seminar

Abstract:

The field of Financial Mathematics is often associated with Monte Carlo simulations, Brownian motions and classical time series models. However, it also embraces a wide variety of applications for machine learning techniques. Within this talk, we would like to present some of our current research considering mortality forecasts as well as agricultural commodities.

Mortality Forecasts

Mortality forecasts are relevant for demographic analysis as well as for the pricing and risk management of insurance contracts. Neural networks and other machine learning techniques have been successfully applied to enhance or surpass existing models. We motivate the use of convolutional neural networks for mortality rate forecasting, evaluate their performance and sketch directions for future research.

Neural networks and other machine learning techniques have been successfully applied to enhance or surpass existing models. We motivate the use of convolutional neural networks for mortality rate forecasting, evaluate their performance and sketch directions for future research.

Agricultural Commodities

In recent time of noticeable climate change, the consideration of external factors, such as weather and economic key figures, becomes even more crucial for a proper valuation of derivatives written on agricultural commodities. Allowing for the inclusion of external factors, a two-step hybrid model using machine learning methods for clustering and classification is proposed. Based on a stochastic price model and a calibration procedure, K-means is performed to identify historical price states. Using forecasts of these price driving factors, this is followed by a price state prediction using K-nearest neighbors and random forest classification.