|
on Big Data |
| By: | Zvezdin Besarabov; Todor Kolev |
| Abstract: | Blockchain technology shows significant results and huge potential for serving as an interweaving fabric that goes through every industry and market, allowing decentralized and secure value exchange, thus connecting our civilization like never before. The standard approach for asset value predictions is based on market analysis with an LSTM neural network. Blockchain technologies, however, give us access to vast amounts of public data, such as the executed transactions and the account balance distribution. We explore whether analyzing this data with modern Deep Leaning techniques results in higher accuracies than the standard approach. During a series of experiments on the Ethereum blockchain, we achieved $4$ times error reduction with blockchain data than an LSTM approach with trade volume data. By utilizing blockchain account distribution histograms, spatial dataset modeling, and a Convolutional architecture, the error was reduced further by 26\%. The proposed methodologies are implemented in an open source cryptocurrency prediction framework, allowing them to be used in other analysis contexts. |
| Date: | 2018–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1810.06696&r=big |
| By: | Milla M\"akinen; Juho Kanniainen; Moncef Gabbouj; Alexandros Iosifidis |
| Abstract: | The existing literature provides evidence that limit order book data can be used to predict short-term price movements in stock markets. This paper proposes a new neural network architecture for predicting return jump arrivals in equity markets with high-frequency limit order book data. This new architecture, based on Convolutional Long Short-Term Memory with Attention, is introduced to apply time series representation learning with memory and to focus the prediction attention on the most important features to improve performance. The data set consists of order book data on five liquid U.S. stocks. The use of the attention mechanism makes it possible to analyze the importance of the inclusion limit order book data and other input variables. By using this mechanism, we provide evidence that the use of limit order book data was found to improve the performance of the proposed model in jump prediction, either clearly or marginally, depending on the underlying stock. This suggests that path-dependence in limit order book markets is a stock specific feature. Moreover, we find that the proposed approach with an attention mechanism outperforms the multi-layer perceptron network as well as the convolutional neural network and Long Short-Term memory model. |
| Date: | 2018–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1810.10845&r=big |
| By: | Johanna Choumert-Nkolo (EDI - Economic Development Initiatives Limited); Pascale Phelinas (CESSMA UMRD 245 - Centre d'études en sciences sociales sur les mondes africains, américains et asiatiques - IRD - Institut de Recherche pour le Développement - Inalco - Institut National des Langues et Civilisations Orientales - UPD7 - Université Paris Diderot - Paris 7, CERDI - Centre d'Études et de Recherches sur le Développement International - Clermont Auvergne - UCA - Université Clermont Auvergne - CNRS - Centre National de la Recherche Scientifique, IRD - Institut de Recherche pour le Développement) |
| Abstract: | Understanding the multiple dimensions of the development process is based on a fundamental need: quality data. This article presents recent progress in household survey protocols, which focus on addressing some of the data collection challenges specific to low- and middle-income countries. Four dimensions of the survey process are explored: sampling, selection of respondents within households, how the questionnaire is administered, and control of measurement errors. Each of these phases has been the subject of significant methodological advances. The first is the contribution of new satellite and computer technologies to sample selection when sampling frames are non-existent or unusable. The second is based on the use of computer media for the administration of questionnaires. The third is the exploration of different interrogation variants using experimental economics methods (recall period, questionnaire administration methods, interrogation strategy, etc.). The fourth is the introduction of new themes related to changes in consumption patterns due to urbanization and labor organization. |
| Abstract: | Comprendre les multiples dimensions du processus de développement repose sur un besoin fondamental : des données de qualité. Cet article présente les progrès récents des protocoles d'enquête auprès des ménages, qui s'attachent à résoudre certaines difficultés de collecte des données spécifiques aux pays à revenu faible et intermédiaire. Quatre dimensions du processus d'enquête sont explorées : l'échantillonnage, la sélection des répondants au sein des ménages, le mode d'administration du questionnaire et le contrôle des erreurs de mesure. Chacune de ces phases a fait l'objet d'avancées méthodologiques importantes. La première est l'apport des nouvelles technologies satellitaires et informatiques à la sélection de l'échantillon lorsque les bases de sondage sont inexistantes ou inutilisables. La seconde repose sur l'utilisation de supports informatiques pour l'administration des questionnaires. La troisième réside dans l'exploration de différentes variantes d'interrogation grâce aux méthodes de l'économie expérimentale (période de rappel, modes d'administration du questionnaire, stratégie d'interrogation etc.). La quatrième correspond à l'introduction de nouvelles thématiques liées aux changements des modes de consommation imputables à l'urbanisation et à l'organisation du travail. |
| Keywords: | Sampling,Questionnaire,Survey bias,Measurement errors.,échantillonnage,biais d’enquête,erreurs de mesure. |
| Date: | 2018–10–05 |
| URL: | http://d.repec.org/n?u=RePEc:hal:wpaper:halshs-01888609&r=big |
| By: | Anttonen, Jetro |
| Abstract: | Abstract In this paper a Bayesian vector autoregressive model for nowcasting the seasonally non-adjusted unemployment rate in EU-countries is developed. On top of the official statistical releases, the model utilizes Google search data and the effect of Google data on the forecasting performance of the model is assessed. The Google data is found to yield modest improvements in forecasting accuracy of the model. To the author’s knowledge, this is the first time the forecasting performance of the Google search data has been studied in the context of Bayesian vector autoregressive model. This paper also adds to the empirical literature on the hyperparameter choice with Bayesian vector autoregressive models. The hyperparameters are set according to the mode of the posterior distribution of the hyperparameters, and this is found to improve the out-of-sample forecasting accuracy of the model significantly, compared to the rule-of-thumb values often used in the literature. |
| Keywords: | Nowcasting, Forecasting, BVAR, Big Data, Unemployment |
| JEL: | C32 C53 C82 E27 |
| Date: | 2018–11–05 |
| URL: | http://d.repec.org/n?u=RePEc:rif:wpaper:62&r=big |
| By: | Farboodi, Maryam; Veldkamp, Laura |
| Abstract: | "Big data" financial technology raises concerns about market inefficiency. A common concern is that the technology might induce traders to extract others' information, rather than produce information themselves. We allow agents to choose how much to learn about future asset values or about others' demands, and explore how improvements in data processing shape these information choices, trading strategies and market outcomes. Our main insight is that unbiased technological change can explain a market-wide shift in data collection and trading strategies. However, in the long run, as data processing technology becomes more and more advanced, both types of data continue to be processed. What keeps the data economy in balance is two competing forces: Data resolves investment risk, but future data creates risk. The efficiency results that follow from these competing forces upend common wisdom. They offer a new take on what makes prices informative and whether trades typically deemed liquidity-providing actually make markets more resilient. |
| Keywords: | Big Data; financial analysis; Fintech; growth; Information Acquisition; liquidity |
| JEL: | E2 G14 |
| Date: | 2018–10 |
| URL: | http://d.repec.org/n?u=RePEc:cpr:ceprdp:13278&r=big |
| By: | Filipski, M.; Belton, B. |
| Abstract: | The rapid growth of fish farming over the past three decades has generated heated debates over the place of aquaculture in rural development. Central to these debates is the question of whether and how aquaculture impacts local incomes and employment, yet little empirical evidence exists on the issue. To address this question, we propose a Local Economy-wide Impact Evaluation (LEWIE) model which nests fish farm models within a general-equilibrium model of their local economy. The model is calibrated using primary data collected from 1102 households in Myanmar s main aquaculture zone, representative of 60% of the country s aquaculture. Using this model, we examine the impact of aquaculture on the incomes and labor market outcomes of fish farming households, but also crop farms and non-farm households in the cluster. Simulating one-acre increases in pond/plot surface we find that: (1) aquaculture generates much higher incomes per-acre than agriculture; (2) aquaculture generates larger income spillovers than agriculture for non-farm households, by way of retail and labor markets; (3) small commercial fish farms generate greater spillovers than large fish farms. These results bolster the notion that fish-farming, notably small-scale commercial aquaculture, may have a significant role to play in rural development and poverty reduction. Acknowledgement : This research was made possible by the support of the United States Agency for International Development (USAID) funded Food Security Policy Project (Associate Award No. AID-482-LA-14-00003), and financial assistance from the Livelihoods and Food Security Trust Fund (LIFT) Grant Support Agreement Number: R 1.4/029/2014 for the project Agrifood Value Chain Development in Myanmar: Implications for Livelihoods of the Rural Poor . We also thank Mekamu Kedir Jamal of the International Food Policy Research Institute for assistance with analysis of satellite imagery and mapping. |
| Keywords: | Resource/Energy Economics and Policy |
| Date: | 2018–07 |
| URL: | http://d.repec.org/n?u=RePEc:ags:iaae18:277461&r=big |