Adaptive Variable Selection for Sequential Prediction in Multivariate Dynamic Models

@article{Lavine2021AdaptiveVS,
  title={Adaptive Variable Selection for Sequential Prediction in Multivariate Dynamic Models},
  author={Isaac Lavine and Michael Lindon and Mike West},
  journal={Bayesian Analysis},
  year={2021}
}
We discuss Bayesian model uncertainty analysis and forecasting in sequential dynamic modeling of multivariate time series. The perspective is that of a decision-maker with a specific forecasting objective that guides thinking about relevant models. Based on formal Bayesian decision-theoretic reasoning, we develop a time-adaptive approach to exploring, weighting, combining and selecting models that differ in terms of predictive variables included. The adaptivity allows for changes in the sets of… 
Bayesian forecasting of multivariate time series: scalability, structure uncertainty and decisions
  • M. West
  • Computer Science
    Annals of the Institute of Statistical Mathematics
  • 2019
TLDR
A main focus is on the “decouple/recouple” concept that enables application of state-space models to increasingly large-scale data, applying to continuous or discrete time series outcomes.
Bayesian Predictive Decision Synthesis
Decision-guided perspectives on model uncertainty expand traditional statistical thinking about managing, comparing and combining inferences from sets of models. Bayesian predictive decision
Bayesian Decision Analysis and Constrained Forecasting
A Bayesian decision analysis perspective on problems of constrained forecasting is presented and developed, motivated by increasing interest in problems of aggregate and hierarchical forecasting
Perspectives on Bayesian Decision Analysis and Constrained Forecasting
Bayesian decision analysis perspectives on problems of constrained forecasting are partly motivated by increasing interest in problems of aggregate and hierarchical forecasting coupled with
Perspectives on Constrained Forecasting
This expository paper discusses Bayesian decision analysis perspectives on problems of constrained forecasting. Foundational and pedagogic discussion contrasts decision analytic approaches with the
Dynamic Ordering Learning in Multivariate Forecasting
In many fields where the main goal is to produce sequential forecasts for decisionmaking problems, the good understanding of the contemporaneous relations among different series is crucial for the
Bayesian Aggregation
  • Yuling Yao
  • Computer Science
    Wiley StatsRef: Statistics Reference Online
  • 2021
TLDR
Among two widely used methods, Bayesian model averaging (BMA) and Bayesian stacking, this article compares their predictive performance, and reviews their theoretical optimality, probabilistic interpretation, practical implementation, and extensions in complex models.
Time Series Momentum Predictability via Dynamic Binary Classification*
Time series momentum strategies are widely applied in the quantitative financial industry and its academic research has grown rapidly since the work of Moskowitz, Ooi, and Pedersen (2012). However,
Trend-Following Strategies via Dynamic Momentum Learning
Time series momentum strategies are widely applied in the quantitative financial industry and its academic research has grown rapidly since the work of Moskowitz, Ooi, and Pedersen (2012). However,
Dynamic Portfolio Allocation in High Dimensions using Sparse Risk Factors
TLDR
This work shows in a dynamic portfolio allocation problem with 452 stocks from the S&P 500 index that the dynamic risk factor model is able to produce more stable and sparse predictions, achieving not just considerable portfolio performance improvements but also higher utility gains for the mean-variance investor.
...
...

References

SHOWING 1-10 OF 65 REFERENCES
Bayesian forecasting of multivariate time series: scalability, structure uncertainty and decisions
  • M. West
  • Computer Science
    Annals of the Institute of Statistical Mathematics
  • 2019
TLDR
A main focus is on the “decouple/recouple” concept that enables application of state-space models to increasingly large-scale data, applying to continuous or discrete time series outcomes.
Multivariate Bayesian Predictive Synthesis in Macroeconomic Forecasting
TLDR
New BPS methodology for a specific subclass of the dynamic multivariate latent factor models implied by BPS theory is developed, motivated by the application context-- sequential forecasting of multiple US macroeconomic time series with forecasts generated from several traditional econometric time series models.
Dynamic Bayesian predictive synthesis in time series forecasting
Bayesian Model Averaging for Linear Regression Models
Abstract We consider the problem of accounting for model uncertainty in linear regression models. Conditioning on a single selected model ignores model uncertainty, and thus leads to the
Dynamic dependence networks: Financial time series forecasting and portfolio decisions
TLDR
Novel state-space models characterizing sparse patterns of dependence among multiple time series extend existing multivariate volatility models to enable scaling to higher numbers of individual time series, and shows how the individual series can be decoupling for sequential analysis and then recoupled for applied forecasting and decision analysis.
Bayesian Analysis of Latent Threshold Dynamic Models
We discuss a general approach to dynamic sparsity modeling in multivariate time series analysis. Time-varying parameters are linked to latent processes that are thresholded to induce zero values
Weighted Bayesian Model Averaging for Portfolio Decisions in Matrix Variate Dynamic Linear Models
In this paper, we assess Bayesian model averaging (BMA) techniques for dynamic linear models (DLMs) with variance matrix discounting. In previous research, the discount factors for the variance
Bayesian Emulation for Multi-Step Optimization in Decision Problems
We discuss the Bayesian emulation approach to computational solution of multi-step portfolio studies in financial time series. "Bayesian emulation for decisions" involves mapping the technical
Bayesian Model Monitoring
TLDR
A simple method of monitoring the predictive performance of a class of Bayesian models is introduced, used as a general diagnostic tool to detect and assess discrepancies between the data and predictions generated from the model.
...
...