Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userPredicting Disaggregated CPI Inflation Components via Hierarchical Recurrent Neural Networks
89
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
We present a hierarchical architecture based on Recurrent Neural Networks (RNNs) for predicting disaggregated inflation components of the Consumer Price Index (CPI). While the majority of existing research is focused mainly on predicting the inflation headline, many economic and financial entities are more interested in its partial disaggregated components. To this end, we developed the novel Hierarchical Recurrent Neural Network (HRNN) model that utilizes information from higher levels in the CPI hierarchy to improve predictions at the more volatile lower levels. Our evaluations, based on a large data-set from the US CPI-U index, indicate that the HRNN model significantly outperforms a vast array of well-known inflation prediction baselines.
rate research
Read More
While researchers increasingly use deep neural networks (DNN) to analyze individual choices, overfitting and interpretability issues remain as obstacles in theory and practice. By using statistical learning theory, this study presents a framework to examine the tradeoff between estimation and approximation errors, and between prediction and interpretation losses. It operationalizes the DNN interpretability in the choice analysis by formulating the metrics of interpretation loss as the difference between true and estimated choice probability functions. This study also uses the statistical learning theory to upper bound the estimation error of both prediction and interpretation losses in DNN, shedding light on why DNN does not have the overfitting issue. Three scenarios are then simulated to compare DNN to binary logit model (BNL). We found that DNN outperforms BNL in terms of both prediction and interpretation for most of the scenarios, and larger sample size unleashes the predictive power of DNN but not BNL. DNN is also used to analyze the choice of trip purposes and travel modes based on the National Household Travel Survey 2017 (NHTS2017) dataset. These experiments indicate that DNN can be used for choice analysis beyond the current practice of demand forecasting because it has the inherent utility interpretation, the flexibility of accommodating various information formats, and the power of automatically learning utility specification. DNN is both more predictive and interpretable than BNL unless the modelers have complete knowledge about the choice task, and the sample size is small. Overall, statistical learning theory can be a foundation for future studies in the non-asymptotic data regime or using high-dimensional statistical models in choice analysis, and the experiments show the feasibility and effectiveness of DNN for its wide applications to policy and behavioral analysis.
Latent dynamics discovery is challenging in extracting complex dynamics from high-dimensional noisy neural data. Many dimensionality reduction methods have been widely adopted to extract low-dimensional, smooth and time-evolving latent trajectories. However, simple state transition structures, linear embedding assumptions, or inflexible inference networks impede the accurate recovery of dynamic portraits. In this paper, we propose a novel latent dynamic model that is capable of capturing nonlinear, non-Markovian, long short-term time-dependent dynamics via recurrent neural networks and tackling complex nonlinear embedding via non-parametric Gaussian process. Due to the complexity and intractability of the model and its inference, we also provide a powerful inference network with bi-directional long short-term memory networks that encode both past and future information into posterior distributions. In the experiment, we show that our model outperforms other state-of-the-art methods in reconstructing insightful latent dynamics from both simulated and experimental neural datasets with either Gaussian or Poisson observations, especially in the low-sample scenario. Our codes and additional materials are available at https://github.com/sheqi/GP-RNN_UAI2019.
We model sectoral production by cascading binary compounding processes. The sequence of processes is discovered in a self-similar hierarchical structure stylized in the economy-wide networks of production. Nested substitution elasticities and Hicks-neutral productivity growth are measured such that the general equilibrium feedbacks between all sectoral unit cost functions replicate the transformation of networks observed as a set of two temporally distant input-output coefficient matrices. We examine this system of unit cost functions to determine how idiosyncratic sectoral productivity shocks propagate into aggregate macroeconomic fluctuations in light of potential network transformation. Additionally, we study how sectoral productivity increments propagate into the dynamic general equilibrium, thereby allowing network transformation and ultimately producing social benefits.
Regulation is commonly viewed as a hindrance to entrepreneurship, but heterogeneity in the effects of regulation is rarely explored. We focus on regional variation in the effects of national-level regulations by developing a theory of hierarchical institutional interdependence. Using the political science theory of market-preserving federalism, we argue that regional economic freedom attenuates the negative influence of national regulation on net job creation. Using U.S. data, we find that regulation destroys jobs on net, but regional economic freedom moderates this effect. In regions with average economic freedom, a one percent increase in regulation results in 14 fewer jobs created on net. However, a standard deviation increase in economic freedom attenuates this relationship by four fewer jobs. Interestingly, this moderation accrues strictly to older firms; regulation usually harms young firm job creation, and economic freedom does not attenuate this relationship.
Since the 1980s, technology business incubators (TBIs), which focus on accelerating businesses through resource sharing, knowledge agglomeration, and technology innovation, have become a booming industry. As such, research on TBIs has gained international attention, most notably in the United States, Europe, Japan, and China. The present study proposes an entrepreneurial ecosystem framework with four key components, i.e., people, technology, capital, and infrastructure, to investigate which factors have an impact on the performance of TBIs. We also empirically examine this framework based on unique, three-year panel survey data from 857 national TBIs across China. We implemented factor analysis and panel regression models on dozens of variables from 857 national TBIs between 2015 and 2017 in all major cities in China and found that a number of factors associated with people, technology, capital, and infrastructure components have various statistically significant impacts on the performance of TBIs at either national model or regional models.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
