|
on Forecasting |
| By: | Souhaib Ben Taieb; Rob J Hyndman |
| Abstract: | Multi-step forecasts can be produced recursively by iterating a one-step model, or directly using a specific model for each horizon. Choosing between these two strategies is not an easy task since it involves a trade-off between bias and estimation variance over the forecast horizon. Using a nonlinear machine learning model makes the tradeoff even more difficult. To address this issue, we propose a new forecasting strategy which boosts traditional recursive linear forecasts with a direct strategy using a boosting autoregression procedure at each horizon. First, we investigate the performance of the proposed strategy in terms of bias and variance decomposition of the error using simulated time series. Then, we evaluate the proposed strategy on real-world time series from two forecasting competitions. Overall, we obtain excellent performance with respect to the standard forecasting strategies. |
| Keywords: | Multi-step forecasting; forecasting strategies; recursive forecasting; direct forecasting; linear time series; nonlinear time series; boosting |
| JEL: | C22 C53 C14 |
| Date: | 2014 |
| URL: | http://d.repec.org/n?u=RePEc:msh:ebswps:2014-13&r=for |
| By: | Wang, Zhu (Federal Reserve Bank of Richmond); Wolman, Alexander L. (Federal Reserve Bank of Richmond) |
| Abstract: | This paper uses transaction-level data from a large discount chain together with zip-code-level explanatory variables to learn about consumer payment choices across size of transaction, location, and time. With three years of data from thousands of stores across the country, we identify important economic and demographic effects; weekly, monthly, and seasonal cycles in payments, as well as time trends and significant state-level variation that is not accounted for by the explanatory variables. We use the estimated model to forecast how the mix of consumer payments will evolve and to forecast future demand for currency. Our estimates based on this large retailer, together with forecasts for the explanatory variables, lead to a benchmark prediction that the cash share of retail sales will decline by 2.54 percentage points per year over the next several years. |
| Keywords: | Payment choice; Money demand; Consumer behavior |
| JEL: | D12 E41 G2 |
| Date: | 2014–04–14 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedrwp:14-09&r=for |
| By: | Grove, Wayne A. (Le Moyne College); Jetter, Michael (Universidad EAFIT) |
| Abstract: | We estimate the relationship between international youth and professional tennis rankings. We find no difference between the predictiveness of rankings from age 14 & Under versus age 16 & Under competitions. The most persistent predictor of professional success is beating older top ranked juniors. Our results reveal stark gender differences. For example, ordinal junior rankings are more strongly associated with professional success for males than for females. In addition, future tennis stars are better signaled by U14 competition outcomes for females, but by U16 results for males. |
| Keywords: | productivity measures, labor supply, career outcomes, tennis |
| JEL: | D82 D91 J16 J22 J24 |
| Date: | 2014–04 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp8103&r=for |
| By: | Pang, Tianxiao; Zhang, Danna; Chong, Terence Tai-Leung |
| Abstract: | This paper examines the asymptotic inference for AR(1) models with a possible structural break in the AR parameter β near the unity at an unknown time k₀. Consider the model y_{t}=β₁y_{t-1}I{t≤k₀}+β₂y_{t-1}I{t>k₀}+ε_{t}, t=1,2,⋯,T, where I{⋅} denotes the indicator function. We examine two cases: Case (I) |β₁| |
| Keywords: | AR(1) model, Change point, Domain of attraction of the normal law, Limiting distribution, Least squares estimator. |
| JEL: | C22 |
| Date: | 2013–12–30 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:55312&r=for |