nep-cmp New Economics Papers
on Computational Economics
Issue of 2016‒12‒18
eight papers chosen by
Stan Miles
Thompson Rivers University

  1. Ensembles of Classifiers for Parallel Categorization of Large Number of Text Documents Expressing Opinions By Frantisek Darena; Jan Zizka
  2. Evaluating the Performance of ANN Prediction System at Shanghai Stock Market in the Period 21-Sep-2016 to 11-Oct-2016 By Barack Wamkaya Wanjawa
  3. Effects of Unilateral Trade Liberalization in South Asian countries: Applications of CGE Models By Selim Raihan
  4. Combinatorial algorithms for the seriation problem By Seminaroti, Matteo
  5. Strategies for Achieving the Sustainable Development Goals (SDGs) in South Asia: Lessons from Policy Simulations By Nagesh Kumar; Matthew Hammill; Selim Raihan; Swayamsiddha Panda
  6. An agent based analysis of the impacts of land use restriction and network structures on participation in conservation reserve programs By Sayed, Iftekhar; John, Tisdell
  7. Updating NAMOD: A Namibian tax-benefit microsimulation model By Gemma Wright; Michael Noble; David McLennan; Michell Mpike
  8. Specification of Spatial-Dynamic Externalities and Implications for Strategic Behavior in Disease Control By Atallah, Shady S.; Gomez, Miguel I.; Conrad, Jon M.

  1. By: Frantisek Darena (Department of Informatics, Faculty of Business and Economics, Mendel Uni- versity in Brno, Zemedelska 1, 613 00 Brno, Czech Republic); Jan Zizka (Department of Informatics, Faculty of Business and Economics, Mendel Uni- versity in Brno, Zemedelska 1, 613 00 Brno, Czech Republic)
    Abstract: Opinions provided by people that used some services or purchased some goods are a rich source of knowledge. The opinion classification, applying mostly supervised classifiers, is one of the essential tasks. Computer’s technological capabilities are still a major obstacle, especially when processing huge volumes of data. This study proposes and evaluates ex- perimentally a parallelism application to the classification of a very large number of contrary opinions expressed as freely written text reviews. Instead of training a single classifier on the entire data set, an ensemble of classifiers is trained on disjunctive subsets of data and a group decision is used for the classification of unlabelled items. The main assessment criteria are computational efficiency and error rates, combined into a single measure to be able to compare ensembles of different sizes. Support vector machines, artificial neural networks, and deci- sion trees, belonging to frequently used classification methods, were examined. The paper demonstrates the suggested method viability when the number of text reviews leads to com- putational complexity, which is beyond the contemporary common PC’s capabilities. Classifi- cation accuracy and the values of other classification performance measures (Precision, Recall, F-measure) did not decrease, which is a positive finding.
    Keywords: text documents, natural language, classification, parallel processing, ensembles of classifiers, machine learning
    JEL: C38 C89
    Date: 2016–12
  2. By: Barack Wamkaya Wanjawa
    Abstract: This research evaluates the performance of an Artificial Neural Network based prediction system that was employed on the Shanghai Stock Exchange for the period 21-Sep-2016 to 11-Oct-2016. It is a follow-up to a previous paper in which the prices were predicted and published before September 21. Stock market price prediction remains an important quest for investors and researchers. This research used an Artificial Intelligence system, being an Artificial Neural Network that is feedforward multi-layer perceptron with error backpropagation for prediction, unlike other methods such as technical, fundamental or time series analysis. While these alternative methods tend to guide on trends and not the exact likely prices, neural networks on the other hand have the ability to predict the real value prices, as was done on this research. Nonetheless, determination of suitable network parameters remains a challenge in neural network design, with this research settling on a configuration of 5:21:21:1 with 80% training data or 4-year of training data as a good enough model for stock prediction, as already determined in a previous research by the author. The comparative results indicate that neural network can predict typical stock market prices with mean absolute percentage errors that are as low as 1.95% over the ten prediction instances that was studied in this research.
    Date: 2016–12
  3. By: Selim Raihan
    Abstract: This paper explores the economy-wide effects of trade liberalization in five South Asian countries (Bangladesh, India, Nepal, Pakistan and Sri Lanka) using updated Social Accounting Matrices (SAM) and the static Computable General Equilibrium (CGE) models of these countries for the year 2012. The CGE framework captures the impact of unilateral trade liberalization on macro-economy, trade, employment and household welfare in the selected countries by tracing the price effects of exogenous shocks, where the variations in prices lead to re-allocation of resources among competing activities, which may alter the factorial income and, hence, the distribution of household income. The results show that trade liberalization measures stimulates growth in employment, for skilled and unskilled labour, as well as real income for all the five South Asian countries. Tariff elimination increases real GDP at factor cost by 3.1 percent in Bangladesh, by 2.5 percent in India, by 2 percent in Nepal, by 0.9 percent in Pakistan, and by 0.6 percent in Sri Lanka. The relative price and wage changes in these five economies are also observed to culminate in a general depreciation of real exchange rates, making their exports more competitive in the world markets.
    Keywords: South Asia, Unilateral Trade Liberalization, Computable General Equilibrium Models, Trade and Employment, Household Income Distribution
    JEL: F14 F16
    Date: 2015–07
  4. By: Seminaroti, Matteo (Tilburg University, School of Economics and Management)
    Abstract: In this thesis we study the seriation problem, a combinatorial problem arising in data analysis, which asks to sequence a set of objects in such a way that similar objects are ordered close to each other. We focus on the combinatorial structure and properties of Robinsonian matrices, a special class of structured matrices which best achieve the seriation goal. Our contribution is both theoretical and practical, with a particular emphasis on algorithms. In Chapter 2 we introduce basic concepts about graphs, permutations and proximity matrices used throughout the thesis. In Chapter 3 we present Robinsonian matrices, discussing their characterizations and recognition algorithms existing in the literature. In Chapter 4 we discuss Lexicographic Breadth-First search (Lex-BFS), a special graph traversal algorithm used in multisweep algorithms for the recognition of several classes of graphs. In Chapter 5 we introduce a new Lex-BFS based algorithm to recognize Robinsonian matrices, which is derived from a new characterization of Robinsonian matrices in terms of straight enumerations of unit interval graphs. In Chapter 6 we introduce the novel Similarity-First Search algorithm (SFS), a weighted version of Lex-BFS which we use in a multisweep algorithm for the recognition of Robinsonian matrices. In Chapter 7 we model the seriation problem as an instance of Quadratic Assignment Problem (QAP) and we show that if the data has a Robinsonian structure, then one can find an optimal solution for QAP using a Robinsonian recognition algorithm. In Chapter 8 we discuss how to solve the seriation problem when the data does not have a Robinsonian structure, by finding a Robinsonian approximation of the original data. Finally, in Chapter 9 we discuss some experiments which we have carried out in order to compare the performance of the algorithms introduced in the thesis.
    Date: 2016
  5. By: Nagesh Kumar (United Nations Economic and Social Commission for Asia and the Pacific (ESCAP) South and South-West Asia Office); Matthew Hammill (United Nations Economic and Social Commission for Asia and the Pacific (ESCAP) South and South-West Asia Office); Selim Raihan; Swayamsiddha Panda (United Nations Economic and Social Commission for Asia and the Pacific (ESCAP) South and South-West Asia Office)
    Abstract: This paper analyzes the major challenges to achieving sustainable development in South Asia as a basis for articulating development strategies for the achievement of the Sustainable Development Goals (SDGs) in the subregion. It identifies key combinations of dimensions of the 2030 Agenda and SDGs that could form core development priorities and maximize interactions for the achievement of the SDGs. The paper further analyzes the policy impacts from select development priorities within a computable general equilibrium framework on economic growth, poverty reduction and employment, among other parameters of development. The results suggest that an industry-oriented structural transformation, enhancing agricultural productivity through sustainable agriculture and overall efficiency improvements through innovations have the potential to lift an additional 71 million people out of poverty, create 56 million additional jobs in South Asia and boost GDP by 15-30 per cent by 2030 over and above the business-as-usual scenario.
    Keywords: South Asia, Sustainable Development Goals (SDGs), Industrialization, Agricultural productivity, Economic growth, Computable General Equilibrium Models
    JEL: C68 O1 O2 O5
    Date: 2016–08
  6. By: Sayed, Iftekhar; John, Tisdell
    Abstract: Conservation covenants are a policy tool for biodiversity and environmental conservation on private lands. They are associated with the withdrawal of development rights by the landholder over a particular piece of land in exchange for financial benefits. Previous studies suggest that existing network structure could influence an individual’s decision to enrol in conservation programs, but there is a lack of comparative analysis of network types to promote more cost-effective results. In this paper, we develop an agent based simulation model to demonstrate the evolution and impact of future land use restrictions on the enrolment of landholders in conservation covenant programs under different network structure. We observe that the nature of the network has a significant impact on program performance. We obtain a lower response to (and higher cost of) conservation covenanting programs when agents are part of a random matching network compared to other networks options. On the other hand, program costs are lower when agents are part of a local uniform matching network. The outcomes indicate that it might be beneficial for the representations to conduct network analysis the project planning stage to fix their programs more attractive to the landowners.
    Keywords: Conservation covenants, Land use restrictions, Multi-agent systems, Simulation, Network analysis, Environmental Economics and Policy, Q24, Q28, D47,
    Date: 2016–12–11
  7. By: Gemma Wright; Michael Noble; David McLennan; Michell Mpike
    Abstract: This paper provides an account of a Nambian tax-benefit microsimulation model—NAMOD—which has been developed for use by government. Following a section on the importance of social security in Namibia and recent related studies, the paper outlines the tax-benefit policies that are included within NAMOD and describes the data challenges and assumptions that had to be made in order to simulate these policies. Results for 2015 are compared with reported administrative data. In spite of current data challenges, NAMOD can be used to help inform social security policy design.
    Keywords: tax-benefit, microsimulation, South Africa
  8. By: Atallah, Shady S.; Gomez, Miguel I.; Conrad, Jon M.
    Abstract: We propose a novel, distance- and density-dependent specification of externalities that captures spatial dynamics within and between neighboring land parcels. We apply the problem to the short- and long-distance diffusion and control of an infectious disease in two privately-owned and ecologically-connected vineyards. Using computational experiments to generate individual and aggregate payoffs, we show how strategic behavior affects diffusion of the disease and the expected present value of the resulting externality. Our results suggest that ignoring the withinparcel spatial dynamics in the model overestimates the social cost of an externality compared to a model that focuses on inter-parcel spatial dynamics only. We find a U-shaped relationship between manager heterogeneity and aggregate payoffs in the presence of an externality, suggesting both positive and negative impacts of increased heterogeneity on strategic behavior and welfare.
    Keywords: Bioeconomic models, Computational methods, Disease control, Grapevine Leafroll Disease, Noncooperative games, Spatial-dynamic externalities, Crop Production/Industries, Research and Development/Tech Change/Emerging Technologies,
    Date: 2016–01

This nep-cmp issue is ©2016 by Stan Miles. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.