SearcharxivSearch

arXiv subjects

Laura Tinsi

Publications and source records attributed to Laura Tinsi.

4 recordsLinked to original sources

Risk bounds for aggregated shallow neural networks using Gaussian prior

Analysing statistical properties of neural networks is a central topic in statistics and machine learning. However, most results in the literature focus on the properties of the neural network minimizing the training error. The goal of this paper is to consider aggregated neural networks using a Gaussian prior. The departure point of our approach is an arbitrary aggregate satisfying the PAC-Bayesian inequality. The main contribution is a precise nonasymptotic assessment of the estimation error appearing in the PAC-Bayes bound. We also review available bounds on the error of approximating a function by a neural network. Combining bounds on estimation and approximation errors, we establish risk bounds that are sharp enough to lead to minimax rates of estimation over Sobolev smoothness classes.

math.ST

Price formation and optimal trading in intraday electricity markets

We develop a tractable equilibrium model for price formation in intraday electricity markets in the presence of intermittent renewable generation. Using stochastic control theory, we identify the optimal strategies of agents with market impact and exhibit the Nash equilibrium in closed form for a finite number of agents as well as in the asymptotic framework of mean field games. Our model reproduces the empirical features of intraday market prices, such as increasing price volatility at the approach of the delivery date and the correlation between price and renewable infeed forecasts, and relates these features with market characteristics like liquidity, number of agents, and imbalance penalty.

q-fin.PR

Decision making with dynamic probabilistic forecasts

We consider a sequential decision making process, such as renewable energy trading or electrical production scheduling, whose outcome depends on the future realization of a random factor, such as a meteorological variable. We assume that the decision maker disposes of a dynamically updated probabilistic forecast (predictive distribution) of the random factor. We propose several stochastic models for the evolution of the probabilistic forecast, and show how these models may be calibrated from ensemble forecasts, commonly provided by weather centers. We then show how these stochastic models can be used to determine optimal decision making strategies depending on the forecast updates. Applications to wind energy trading are given.

q-fin.TR

Price formation and optimal trading in intraday electricity markets with a major player

We study price formation in intraday electricity markets in the presence of intermittent renewable generation. We consider the setting where a major producer may interact strategically with a large number of small producers. Using stochastic control theory we identify the optimal strategies of agents with market impact and exhibit the Nash equilibrium in closed form in the asymptotic framework of mean field games with a major player. This is a companion paper to [Féron, Tankov, and Tinsi, Price formation and optimal trading in intraday electricity markets, arXiv:2009.04786, 2020], where a similar model is developed in the setting of identical agents.

q-fin.PR