We propose a near explosive random coefficient autoregressive model (NERC) to obtain predictive probabilities of the apparition and devolution of bubbles. The distribution of the autoregressive coefficient of this model is allowed to be centred at an O(T−α) distance of unity, with α ∈ (0, 1). When the expectation of the autoregressive coefficient lies on the explosive side of unity, the NERC helps to model the temporary explosiveness of time series and obtain related predictive probabilities. We study the asymptotic properties of the NERC and provide a procedure for inference on the parameters. In empirical illustrations, we estimate predictive probabilities of bubbles or flash crashes in financial asset prices.
[Please read the research paper here]
A data-cloning SMC2 algorithm is proposed as a general-purpose, global optimization routine for the maximum likelihood estimation of models with latent variables. In the SMC2 phase, the method first marginalizes out the latent variable(s) by applying one layer of SMC at a fixed parameter value and then searches for the optimal parameters through another layer of SMC. The data-cloning phase is deployed to ensure global convergence by dampening multi-modality and to reduce the Monte Carlo error associated with SMC. This new method has broad applicability and is massively parallelizable through leveraging modern multi-core CPU or GPU computing.
[Please read the research paper here]
This article compares a general cross-lagged model (GCLM) to other panel data methods based on their coherence with a causal logic and pragmatic concerns regarding modeled dynamics and hypothesis testing. We examine three “static” models that do not incorporate temporal dynamics: random- and fixed-effects models that estimate contemporaneous relationships; and latent curve models. We then describe “dynamic” models that incorporate temporal dynamics in the form of lagged effects: cross-lagged models estimated in a structural equation model (SEM) or multilevel model (MLM) framework; Arellano-Bond dynamic panel data methods; and autoregressive latent trajectory models. We describe the implications of overlooking temporal dynamics in static models and show how even popular cross-lagged models fail to control for stable factors over time. We also show that Arellano-Bond and autoregressive latent trajectory models have various shortcomings. By contrasting these approaches, we clarify the benefits and drawbacks of common methods for modeling panel data, including the GCLM approach we propose. We conclude with a discussion of issues regarding causal inference, including difficulties in separating different types of time-invariant and time-varying effects over time.
[Please read the research paper here]
This is the first paper in a series of two that synthesizes, compares, and extends methods for causal inference with longitudinal panel data in a structural equation modeling (SEM) framework. Starting with a cross-lagged approach, this paper builds a general cross-lagged panel model (GCLM) with parameters to account for stable factors while increasing the range of dynamic processes that can be modeled. We illustrate the GCLM by examining the relationship between national income and subjective well-being (SWB), showing how to examine hypotheses about short-run (via Granger-Sims tests) versus long-run effects (via impulse responses). When controlling for stable factors, we find no short-run or long-run effects among these variables, showing national SWB to be relatively stable, whereas income is less so. Our second paper addresses the differences between the GCLM and other methods.
[Please read the research paper here]
The dynamic traveling salesman problem with stochastic release dates (DTSP-srd) is a problem in which a supplier has to deliver parcels to its customers. These parcels are delivered to its depot while the distribution is taking place. The arrival time of a parcel to the depot is called its release date. In the DTSP-srd, release dates are stochastic and dynamically updated as the distribution takes place. The objective of the problem is the minimization of the total time needed to serve all customers, given by the sum of the traveling time and the waiting time at the depot. The problem is represented as a Markov Decision Process and is solved through a reoptimization approach. Two models are proposed for the problem to be solved at each stage. The first model is stochastic and exploits the entire probabilistic information available for the release dates. The second model is deterministic and uses an estimation of the release dates. An instance generation procedure is proposed to simulate the evolution of the information to perform computational tests. The results show that a more frequent reoptimization provides better results across all tested instances and that the stochastic model performs better than the deterministic model. The main drawback of the stochastic model lies in the computational time required to evaluate a solution, which makes an iteration of the heuristic substantially more time-consuming than in the case where the deterministic model is used.
[Please read the research paper here]
This paper addresses the Inventory Routing Problem with Pickups and Deliveries (IRP-PD). A single commodity has to be picked up from several origins and distributed to several destinations. The commodity is made available at the supplier depot and at the pickup customers, with known amounts in each period of a discretized time horizon. The commodity is consumed by the delivery customers, whose demand, in each period of the time horizon, is known. The vehicles start and end their routes at the supplier depot. The objective is to determine a collection and distribution plan minimizing the sum of routing and inventory costs, satisfying inventory and capacity constraints. A mixed integer linear programming model is presented for the IRP-PD. Valid inequalities are proposed that are adapted from the inventory routing, the vehicle routing and the lot-sizing problems literature. Moreover, a new set of valid inequalities, called interval inequalities, is presented. A branch-and-cut algorithm is designed and tested on a large set of instances. Computational results show that the proposed approach is able to solve to optimality 946 out of 1280 instances. Moreover, the approach outperforms the state-of-the art method for the single vehicle case, increasing the number of instances solved to optimality from 473 to 538 out of 640 instances tested, with 133 new best known solutions.
[Please read the research paper here]
Long-run restrictions are a very popular method for identifying structural vector autoregressions, but they suffer from weak identification when the data is very persistent, i.e., when the highest autoregressive roots are near unity. Near unit roots introduce additional nuisance parameters and make standard weak-instrument robust methods of inference inapplicable. We develop a method of inference that is robust to both weak identification and strong persistence. The method is based on a combination of the Anderson-Rubin test with instruments derived by filtering potentially non-stationary variables to make them near stationary using the IVX instrumentation method of Magdalinos and Phillips (2009). We apply our method to obtain robust confidence bands on impulse responses in two leading applications in the literature.
[Please read the research paper here]
Following widespread availability of computerized databases, much research has correlated bibliometric measures from papers or patents to subsequent success, typically measured as the number of publications or citations. Building on this large body of work, we ask the following questions: given available bibliometric information in one year, along with the combined theories on sources of creative breakthroughs from the literatures on creativity and innovation, how accurately can we explain the impact of authors in a given research community in the following year? In particular, who is most likely to publish, publish highly cited work, and even publish a highly cited outlier? And, how accurately can these existing theories predict breakthroughs using only contemporaneous data? After reviewing and synthesizing (often competing) theories from the literatures, we simultaneously model the collective hypotheses based on available data in the year before RNA interference was discovered. We operationalize author impact using publication count, forward citations, and the more stringent definition of being in the top decile of the citation distribution. Explanatory power of current theories altogether ranges from less than 9% for being top cited to 24% for productivity. Machine learning (ML) methods yield similar findings as the explanatory linear models, and tangible improvement only for non-linear Support Vector Machine models. We also perform predictions using only existing data until 1997, and find lower predictability than using explanatory models. We conclude with an agenda for future progress in the bibliometric study of creativity and look forward to ML research that can explain its models.
[Please read the research paper here]
We study the directed network design problem with relays (DNDPR) whose aim is to construct a minimum cost network that enables the communication of a given set of origin-destination pairs. Thereby, expensive signal regeneration devices need to be placed to cover communication distances exceeding a predefined threshold. Applications of the DNDPR arise in telecommunications and transportation. We propose two new integer programming formulations for the DNDPR. The first one is a flow-based formulation with a pseudo-polynomial number of variables and constraints and the second is a cut-based formulation with an exponential number of constraints. Fractional distance values are handled efficiently by augmenting both models with an exponentially-sized set of infeasible path constraints. We develop branch-and-cut algorithms and also consider valid inequalities to strengthen the obtained dual bounds and to speed up convergence. The results of our extensive computational study on diverse sets of benchmark instances show that our algorithms outperform the previous state-of-the-art method based on column generation.
[Please read the research paper here]
This paper introduces the (α,Γ) -descent, an iterative algorithm which operates on measures and performs α-divergence minimisation in a Bayesian framework. This gradient-based procedure extends the commonly-used variational approximation by adding a prior on the variational parameters in the form of a measure. We prove that for a rich family of functions Γ, this algorithm leads at each step to a systematic decrease in the α-divergence and derive convergence results. Our framework recovers the Entropic Mirror Descent algorithm and provides an alternative algorithm that we call the Power Descent. Moreover, in its stochastic formulation, the (α,Γ)-descent allows to optimise the mixture weights of any given mixture model without any information on the underlying distribution of the variational parameters. This renders our method compatible with many choices of parameters updates and applicable to a wide range of Machine Learning tasks. We demonstrate empirically on both toy and real-world examples the benefit of using the Power Descent and going beyond the Entropic Mirror Descent framework, which fails as the dimension grows.