Search arXivSearch

arXiv subjects

Oliver Grothe

Publications and source records attributed to Oliver Grothe.

12 recordsLinked to original sources

Learning Nonparametric High-Dimensional Generative Models: The Empirical-Beta-Copula Autoencoder

By sampling from the latent space of an autoencoder and decoding the latent space samples to the original data space, any autoencoder can simply be turned into a generative model. For this to work, it is necessary to model the autoencoder's latent space with a distribution from which samples can be obtained. Several simple possibilities (kernel density estimates, Gaussian distribution) and more sophisticated ones (Gaussian mixture models, copula models, normalization flows) can be thought of and have been tried recently. This study aims to discuss, assess, and compare various techniques that can be used to capture the latent space so that an autoencoder can become a generative model while striving for simplicity. Among them, a new copula-based method, the Empirical Beta Copula Autoencoder, is considered. Furthermore, we provide insights into further aspects of these methods, such as targeted sampling or synthesizing new data with specific features.

stat.ML

A hybrid model for day-ahead electricity price forecasting: Combining fundamental and stochastic modelling

The accurate prediction of short-term electricity prices is vital for effective trading strategies, power plant scheduling, profit maximisation and efficient system operation. However, uncertainties in supply and demand make such predictions challenging. We propose a hybrid model that combines a techno-economic energy system model with stochastic models to address this challenge. The techno-economic model in our hybrid approach provides a deep understanding of the market. It captures the underlying factors and their impacts on electricity prices, which is impossible with statistical models alone. The statistical models incorporate non-techno-economic aspects, such as the expectations and speculative behaviour of market participants, through the interpretation of prices. The hybrid model generates both conventional point predictions and probabilistic forecasts, providing a comprehensive understanding of the market landscape. Probabilistic forecasts are particularly valuable because they account for market uncertainty, facilitating informed decision-making and risk management. Our model delivers state-of-the-art results, helping market participants to make informed decisions and operate their systems more efficiently.

econ.EM

Enhancing Energy System Models Using Better Load Forecasts

Energy system models require a large amount of technical and economic data, the quality of which significantly influences the reliability of the results. Some of the variables on the important data source ENTSO-E transparency platform, such as transmission system operators' day-ahead load forecasts, are known to be biased. These biases and high errors affect the quality of energy system models. We propose a simple time series model that does not require any input variables other than the load forecast history to significantly improve the transmission system operators' load forecast data on the ENTSO-E transparency platform in real-time, i.e., we successively improve each incoming data point. We further present an energy system model developed specifically for the short-term day-ahead market. We show that the improved load data as inputs reduce pricing errors of the model, with strong reductions particularly in times when prices are high and the market is tight.

econ.GN

From point forecasts to multivariate probabilistic forecasts: The Schaake shuffle for day-ahead electricity price forecasting

Modeling price risks is crucial for economic decision making in energy markets. Besides the risk of a single price, the dependence structure of multiple prices is often relevant. We therefore propose a generic and easy-to-implement method for creating multivariate probabilistic forecasts based on univariate point forecasts of day-ahead electricity prices. While each univariate point forecast refers to one of the day's 24 hours, the multivariate forecast distribution models dependencies across hours. The proposed method is based on simple copula techniques and an optional time series component. We illustrate the method for five benchmark data sets recently provided by Lago et al. (2020). Furthermore, we demonstrate an example for constructing realistic prediction intervals for the weighted sum of consecutive electricity prices, as, e.g., needed for pricing individual load profiles.

econ.EM

Signing the Supermask: Keep, Hide, Invert

The exponential growth in numbers of parameters of neural networks over the past years has been accompanied by an increase in performance across several fields. However, due to their sheer size, the networks not only became difficult to interpret but also problematic to train and use in real-world applications, since hardware requirements increased accordingly. Tackling both issues, we present a novel approach that either drops a neural network's initial weights or inverts their respective sign. Put simply, a network is trained by weight selection and inversion without changing their absolute values. Our contribution extends previous work on masking by additionally sign-inverting the initial weights and follows the findings of the Lottery Ticket Hypothesis. Through this extension and adaptations of initialization methods, we achieve a pruning rate of up to 99%, while still matching or exceeding the performance of various baseline and previous models. Our approach has two main advantages. First, and most notable, signed Supermask models drastically simplify a model's structure, while still performing well on given tasks. Second, by reducing the neural network to its very foundation, we gain insights into which weights matter for performance. The code is available on GitHub.

cs.LG

Smooth bootstrapping of copula functionals

The smooth bootstrap for estimating copula functionals in small samples is investigated. It can be used both to gauge the distribution of the estimator in question and to augment the data. Issues arising from kernel density and distribution estimation in the copula domain are addressed, such as how to avoid the bounded domain, which bandwidth matrix to choose, and how the smoothing can be carried out. Furthermore, we investigate how the smooth bootstrap impacts the underlying dependence structure or the functionals in question and under which conditions it does not. We provide specific examples and simulations that highlight advantages and caveats of the approach.

stat.CO

A multivariate extension of the Lorenz curve based on copulas and a related multivariate Gini coefficient

We propose an extension of the univariate Lorenz curve and of the Gini coefficient to the multivariate case, i.e., to simultaneously measure inequality in more than one variable. Our extensions are based on copulas and measure inequality stemming from inequality in every single variable as well as inequality stemming from the dependence structure of the variables. We derive simple nonparametric estimators for both instruments and apply them exemplary to data of individual income and wealth for various countries.

stat.ME

The Gibbs Sampler with Particle Efficient Importance Sampling for State-Space Models

We consider Particle Gibbs (PG) as a tool for Bayesian analysis of non-linear non-Gaussian state-space models. PG is a Monte Carlo (MC) approximation of the standard Gibbs procedure which uses sequential MC (SMC) importance sampling inside the Gibbs procedure to update the latent and potentially high-dimensional state trajectories. We propose to combine PG with a generic and easily implementable SMC approach known as Particle Efficient Importance Sampling (PEIS). By using SMC importance sampling densities which are approximately fully globally adapted to the targeted density of the states, PEIS can substantially improve the mixing and the efficiency of the PG draws from the posterior of the states and the parameters relative to existing PG implementations. The efficiency gains achieved by PEIS are illustrated in PG applications to a univariate stochastic volatility model for asset returns, a non-Gaussian nonlinear local-level model for interest rates, and a multivariate stochastic volatility model for the realized covariance matrix of asset returns.

stat.CO

A higher order correlation unscented Kalman filter

Many nonlinear extensions of the Kalman filter, e.g., the extended and the unscented Kalman filter, reduce the state densities to Gaussian densities. This approximation gives sufficient results in many cases. However, this filters only estimate states that are correlated with the observation. Therefore, sequential estimation of diffusion parameters, e.g., volatility, which are not correlated with the observations is not possible. While other filters overcome this problem with simulations, we extend the measurement update of the Gaussian two-moment filters by a higher order correlation measurement update. We explicitly state formulas for a higher order unscented Kalman filter within a continuous-discrete state space. We demonstrate the filter in the context of parameter estimation of an Ornstein-Uhlenbeck process.

stat.ME

Vine Constructions of Levy Copulas

Levy copulas are the most general concept to capture jump dependence in multivariate Levy processes. They translate the intuition and many features of the copula concept into a time series setting. A challenge faced by both, distributional and Levy copulas, is to find flexible but still applicable models for higher dimensions. To overcome this problem, the concept of pair copula constructions has been successfully applied to distributional copulas. In this paper, we develop the pair construction for Levy copulas (PLCC). Similar to pair constructions of distributional copulas, the pair construction of a d-dimensional Levy copula consists of d(d-1)/2 bivariate dependence functions. We show that only d-1 of these bivariate functions are Levy copulas, whereas the remaining functions are distributional copulas. Since there are no restrictions concerning the choice of the copulas, the proposed pair construction adds the desired flexibility to Levy copula models. We discuss estimation and simulation in detail and apply the pair construction in a simulation study.

stat.OT

Measuring Association between Random Vectors

This paper suggests five measures of association between two random vectors X = (X_1, ..., X_p) and Y = (Y_1, ..., Y_q). They are copula based and therefore invariant with respect to the marginal distributions of the components X_i and Y_j. The measures capture positive as well as negative association of X and Y. In case p = q = 1 they reduce to Spearman's rho. Various properties of these new measures are investigated. Nonparametric estimators, based on ranks, for the measures are derived and their small sample behaviour is investigated by simulation. The measures are applied to characterise strength and direction of association of bond and stock indices of five countries over time.

stat.ME

Estimating correlation and covariance matrices by weighting of market similarity

We discuss a weighted estimation of correlation and covariance matrices from historical financial data. To this end, we introduce a weighting scheme that accounts for similarity of previous market conditions to the present one. The resulting estimators are less biased and show lower variance than either unweighted or exponentially weighted estimators. The weighting scheme is based on a similarity measure which compares the current correlation structure of the market to the structures at past times. Similarity is then measured by the matrix 2-norm of the difference of probe correlation matrices estimated for two different times. The method is validated in a simulation study and tested empirically in the context of mean-variance portfolio optimization. In the latter case we find an enhanced realized portfolio return as well as a reduced portfolio volatility compared to alternative approaches based on different strategies and estimators.

q-fin.ST