1,721,004 research outputs found
Generalised autocovariances and spectral estimators
No full textA class of models for the time-varying spectrum of a locally stationary process is introduced. The models are specified in the frequency domain and
the class depends on a power parameter that applies to the spectrum so that it can be locally represented by a finite Fourier polynomial. The coefficients of the polynomial are dynamic generalised cepstral coefficients that have an interpretation as generalised autocovariances. The dynamics of
the generalised cepstral coefficients are determined according to a linear combination of logistic transition functions of the time index. Estimation
is carried out in the frequency domain based on the generalised Whittle likelihood
Stationarity of a general class of observation driven models for discrete valued processes
No full textA large variety of time series observation-driven models for binary and count data are currently used in different contexts. Despite the importance of station- arity and ergodicity to ensure suitable results, for many of these models stationarity is not yet proved. We specify a general class of observation-driven models for dis- crete valued processes, which encompasses the most frequently used models. Then, we show strict stationarity by means of Feller properties and establish easy-to-check stationarity conditions
Testing the Hypothesis of Enhanced Design in Fast Fashion Industry using Internet as a Source of Data
No full text
Inference with the Unscented Kalman Filter and optimization of sigma points
No full textWe investigate the accuracy of inference in a chaotic dynamical system (Duffing oscillator) with the Unscented Kalman Filter and quantify the dependence on the sample size and the signal to noise ratio. In order to improve convergence to the true parameters in the case of a bad initialization of the algorithm, we optimize the location of sigma points with Bayesian optimisation
Inference with the Unscented Kalman Filter and optimization of sigma points for the Duffing process
No full textText Based Pricing Modelling: an Application to the Fashion Industry
No full textBy using internet as a source of data, we estimate pricing models based on the information contained in the description of items on sale. The novel application is on one category of the Fashion industry. Our estimation strategy uses text mining and methods of sparse modelling, namely shrinkage methods and dimension reduction methods, with the aim of obtaining a model which gives the best out-of-sample predictive performance with a high level of interpretability.
The results show that compared with the simple predictor, the average price, the models developed in the paper produce a decrease in the pricing error which is up to 7.7% when the brand is considered and up to 58.3%, when the brand is not considered, in the case when shrinkage methods are used. When dimension reduction methods are used, the decrease is up to 41.4% when brand is included and up to 66.8% in the no-brand case
A Robust Score-Driven Filter for Multivariate Time Series
Full text linkA multivariate score-driven filter is developed to extract signals from noisy
vector processes. By assuming that the conditional location vector from a
multivariate Student's t distribution changes over time, we construct a robust
filter which is able to overcome several issues that naturally arise when
modeling heavy-tailed phenomena and, more in general, vectors of dependent
non-Gaussian time series. We derive conditions for stationarity and
invertibility and estimate the unknown parameters by maximum likelihood (ML).
Strong consistency and asymptotic normality of the estimator are proved and the
finite sample properties are illustrated by a Monte-Carlo study. From a
computational point of view, analytical formulae are derived, which consent to
develop estimation procedures based on the Fisher scoring method. The theory is
supported by a novel empirical illustration that shows how the model can be
effectively applied to estimate consumer prices from home scanner data
Reinforcement learning in modern biostatistics: benefits, challenges and new proposals
Full text linkApplications of reinforcement learning (RL) for supporting, managing and improving decision-making are becoming increasingly popular in a variety of medicine and healthcare domains where the problem has a sequential nature. By continuously interacting with the underlying environment, RL techniques are able to learn by trial-and-error on how to take better actions in order to maximize an outcome of interest over time. However, if on one hand RL offers a new powerful framework, on the other hand it poses some unique challenges for data analysis and interpretability, which call for new statistical techniques in both predictive and descriptive learning.
Notably, several methodological challenges, for which the contribution of the biostatistical community may play a crucial role, limit the use of RL in real life. In an aim to bridge the statistics and RL communities, we start by assimilating the different existing RL terminologies, notations and approaches into a coherent body of work, and by translating them from a machine learning (ML) to a statistical perspective. Then, through a comprehensive methodological review, we report and discuss the state-of-the-art RL-based research in healthcare. Two main applied domains emerged: 1) adaptive interventions (AIs), encompassing both dynamic treatment regimes and just-in-time adaptive interventions in mobile health (mHealth); and 2) adaptive designs of clinical trials, specifically dose-finding designs and adaptive randomization. We illustrate existing RL-based methods in these areas, discussing their benefits and existing open problems that may impact their application in real life. A major barrier to adopting RL in real-world experiments is the lack of clarity on how statistical analyses and inference are impacted. In clinical trials for example, if on one side, to achieve the practical (and more ethical) goal of improving patients’ benefits, RL may have better abilities in terms of maximising clinical outcomes by adaptively randomizing participants to the best evidence-based treatment; on the other side, to achieve the scientific goal of e.g., discovering whether one treatment is more effective compared to a control treatment, less is known about their inferential properties. Through a simulation study, we investigate the challenges of conducting hypothesis testing from data collected through a class of RL, i.e., multi-armed bandits (MABs), outlining the harms MAB algorithms can cause to traditional statistical tests’ type-I error and power. This empirical evaluation provides guidance to two alternative ways of pursuing improved statistical hypothesis testing: 1) to explore ways of modifying the test statistic using knowledge of the adaptive data collection nature; 2) to modify the algorithm or framework for a more sensitive problem to both statistical inference as well as reward maximization. Focusing on the Thompson Sampling (a randomized MAB strategy), we show how a modified version of it results in an optimal intermediate between these two objectives. These findings can provide insights into how challenges can be surmounted by bridging machine learning, statistics, and applied sciences, to conduct adaptive experiments in the real-world, aiming to simultaneously help individuals and advance
scientific research. We finally combine our methodological knowledge with a motivating mHealth study for improving physical activity, to illustrate the tremendous collaboration opportunities between statistics and RL researchers in the space of developing adaptive interventions into the increasingly growing area of mHealth
- …
