This paper presents a Long Short Term Memory Recurrent Neural Network and Hidden Markov Model (LSTM-HMM) to predict China’s Gross Domestic Product (GDP) fluctuation state within a rolling time ...window. We compare the predictive power of LSTM-HMM with other dynamic forecast systems within different time windows, which involves the Hidden Markov Model (HMM), Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) and LSTM-HMM with an input of monthly Consumer Price Index (CPI) or quarterly CPI within 4-year, 6-year, 8-year and 10-year time window. These forecasting models employed in our empirical analysis share the basic HMM structure but differ in the generation of observable CPI fluctuation states. Our forecasting results suggest that (1) among all the models, LSTM-HMM generally performs better than the other models; (2) the model performance can be improved when model input transforms from quarterly to monthly; (3) among all the time windows, models within 10-year time window have better overall performance; (4) within 10-year time window, the LSTM-HMM, with either quarterly or monthly input, has the best accuracy and consistency.
This article proposes a combination model, which is composed of latent Dirichlet allocation model, TF-IDF feature extraction algorithm and Euclidean distance measurement method, to identify and judge ...whether the similarities between multiple policy texts exist or not. With the help of actual data result, this will drive the relevant government agencies to figure out problems in a timely manner and provide a decision-making basis for them to formulate and optimise appropriate economic policies. To this end, this article analyses and studies the four types of economic texts that are classified as Insurance, Banking, Tax and Finance from the Central Government of Hebei province and Shijiazhuang city levels. Also, we consider Beijing, Shanghai and Guangdong. Experimental results show that (1) the combination model can quickly and effectively recognise and determine whether there are similarities between multiple economic policy texts; (2) similarities exist or not between the central, provincial and municipal level policy texts depending on the comparison of the distance values across them; (3) the smaller the distance value between economic policy texts of the same kind, the higher the similarity in them; and (4) the distance values between the six policy texts in Finance, Insurance, Bank and Tax categories are ranked from low to high. In terms of similarity, the Finance category is the highest, followed by Insurance and Bank, and the Tax category is the lowest.
We present a new Hawkes-Contact model that combines a Hawkes process and a finite-range contact process to model the stock price movements, especially under the impact of news and other information ...flows that could lead to contagious effects. To fully capture the underlying price process, we take the Hawkes process to track the full pathway of historical prices on their future movements and the contact process to capture the impact from news/investment sentiment. We compare this full model to a univariate Hawkes process that works as a benchmark model through analyzing their statistical properties using both simulated returns and the real 5-min returns of the crude oil index (Wind CZCE-TA). The statistical properties include probability density function, complementary cumulative distribution function, and Lempel-Ziv Complex. Our results show that the real returns' distribution is often far from normal, but the simulated returns through the Hawkes or Hawke-Contact model can achieve close fit to the real returns and exhibit similar statistical properties. More importantly, the Hawkes-Contact model performs better than the simple Hawkes model in capturing characteristics in the return movements, which indicates that the price evolution is also driven by the news sentiment created after them.
This study aims to comprehensively review a recently emerging multidisciplinary area related to the application of deep learning methods in cryptocurrency research. We first review popular deep ...learning models employed in multiple financial application scenarios, including convolutional neural networks, recurrent neural networks, deep belief networks, and deep reinforcement learning. We also give an overview of cryptocurrencies by outlining the cryptocurrency history and discussing primary representative currencies. Based on the reviewed deep learning methods and cryptocurrencies, we conduct a literature review on deep learning methods in cryptocurrency research across various modeling tasks, including price prediction, portfolio construction, bubble analysis, abnormal trading, trading regulations and initial coin offering in cryptocurrency. Moreover, we discuss and evaluate the reviewed studies from perspectives of modeling approaches, empirical data, experiment results and specific innovations. Finally, we conclude this literature review by informing future research directions and foci for deep learning in cryptocurrency.
Display omitted
Artificial intelligence; Machine learning; Social sciences; Economics
This paper presents a Long Short Term Memory Recurrent Neural Network and Hidden Markov Model (LSTM-HMM) to predict China’s Gross Domestic Product (GDP) fluctuation state within a rolling time ...window. We compare the predictive power of LSTM-HMM with other dynamic forecast systems within different time windows, which involves the Hidden Markov Model (HMM), Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) and LSTM-HMM with an input of monthly Consumer Price Index (CPI) or quarterly CPI within 4-year, 6-year, 8-year and 10-year time window. These forecasting models employed in our empirical analysis share the basic HMM structure but differ in the generation of observable CPI fluctuation states. Our forecasting results suggest that (1) among all the models, LSTM-HMM generally performs better than the other models; (2) the model performance can be improved when model input transforms from quarterly to monthly; (3) among all the time windows, models within 10-year time window have better overall performance; (4) within 10-year time window, the LSTM-HMM, with either quarterly or monthly input, has the best accuracy and consistency.
Applying the theory of statistical physics systems – the voter model, a random stock price model is modeled and studied in this paper, where the voter model is a continuous time Markov process. In ...this price model, for the different parameters values of the intensity
λ, the lattice dimension
d, the initial density
θ, and the multivariate set (
θ,
λ), we discuss and analyze the statistical behaviors of the price model. Moreover, we investigate the power-law distributions, the long-term memory of returns and the volatility clustering phenomena for the Chinese stock indices. The database is from the indices of Shanghai and Shenzhen in the 6-year period from July 2002 to June 2008. Further, the comparisons of the empirical research and the simulation data are given.
Large space buildings often have high HVAC energy consumption, and the thermal pressure ventilation can be utilized to reduce the using of chiller. Temperatures were tested on two naturally ...ventilated large spaces, and it increased linearly towards the ceiling with its gradients α being in range of 0.1-0.4 °C/m. The traditional stack effect model is modified by introducing thermal stratification. For three hot and humid cities in China, the acceptable indoor air temperatures are discussed and determined based on its monthly outdoor air temperatures. Investigations are done to find out parameters' influences on the indoor thermal environment. It is known that both the dimensionless neutral plane height H
n
/H and the volumetric flow rate per unit floor area l are not sensitive to outdoor climate, but decrease with increasing of the lower-upper opening area ratio R
ab
; the occupied air temperature t
n
increases with the increase of R
ab
or q but with the decrease of α or H; under case of 0.2 °C/m and 150 W/m
2
, Nanjing has the maximum scopes of available upper opening areas in transition season to achieve natural ventilation cooling potential. Such results would be useful in design and management of upper openings for large spaces.
