Computational Economics
http://lists.repec.org/mailman/listinfo/nep-cmp
Computational Economics
2019-06-24
Online Block Layer Decomposition schemes for training Deep Neural Networks
http://d.repec.org/n?u=RePEc:aeg:report:2019-06&r=cmp
Deep Feedforward Neural Networks' (DFNNs) weights estimation relies on the solution of a very large nonconvex optimization problem that may have many local (no global) minimizers, saddle points and large plateaus. Furthermore, the time needed to find good solutions to the training problem heavily depends on both the number of samples and the number of weights (variables). In this work, we show how Block Coordinate Descent (BCD) methods can be applied to improve the performance of state-of-the-art algorithms by avoiding bad stationary points and flat regions. We first describe a batch BCD method able to effectively tackle difficulties due to the network's depth; then we further extend the algorithm proposing an online BCD scheme able to scale with respect to both the number of variables and the number of samples. We perform extensive numerical results on standard datasets using different deep networks, and we showed how the application of (online) BCD methods to the training phase of DFNNs permits to outperform standard batch/online algorithms leading to an improvement on both the training phase and the generalization performance of the networks.
Laura Palagi
Ruggiero Seccia
Deep Feedforward Neural Networks ; Block coordinate decomposition ; Online Optimization ; Large scale optimization
2019
An agent-based model for designing a financial market that works well
http://d.repec.org/n?u=RePEc:arx:papers:1906.06000&r=cmp
Designing a financial market that works well is very important for developing and maintaining an advanced economy, but is not easy because changing detailed rules, even ones that seem trivial, sometimes causes unexpected large impacts and side effects. A computer simulation using an agent-based model can directly treat and clearly explain such complex systems where micro processes and macro phenomena interact. Many effective agent-based models investigating human behavior have already been developed. Recently, an artificial market model, which is an agent-based model for a financial market, has started to contribute to discussions on rules and regulations of actual financial markets. I introduce an artificial market model to design financial markets that work well and describe a previous study investigating tick size reduction. I hope that more artificial market models will contribute to designing financial markets that work well to further develop and maintain advanced economies.
Takanobu Mizuta
2019-06
Branching with Hyperplanes in the Criterion Space:the Frontier Partitioner Algorithm for Biobjective Integer Programming
http://d.repec.org/n?u=RePEc:aeg:report:2019-03&r=cmp
We present an algorithm for finding the complete Pareto frontier of biobjective integer programming problems. The method is based on the solution of a finite number of integer programs, each of them returning a Pareto optimal point. The feasible sets of the integer programs are built from the original feasible set, by adding cuts that separate efficient solutions. Providing the existence of an oracle to solve suitably defined single objective integer subproblems, the algorithm can handle biobjective nonlinear integer problems, in particular biobjective convex quadratic integer optimization problems. Our numerical experience on a benchmark of biobjective integer linear programming instances shows the efficiency of the approach in comparison with existing state-of-the-art methods. Further experiments on biobjective integer quadratic programming instances are reported.
Marianna De Santis
Giorgio Grani
Laura Palagi
Multiobjective Optimization ; Integer Programming ; Criterion Space Search
2019
An assessment of IPA public investment in R&I in Albania
http://d.repec.org/n?u=RePEc:ipt:iptwpa:jrc115438&r=cmp
The European Commission's Joint Research Centre (JRC) is supporting an Innovation Agenda for the Western Balkans (Albania, Bosnia and Herzegovina, Kosovo, Montenegro, Republic of North Macedonia and Serbia). Smart Specialisation is the European Union (EU) place-based policy aiming at more thematic concentration in research and innovation (R&I) investments via the evidence-based identification of the strengths and potential of a given economy. Access to data and economic analysis are key to a better identification of both current and future socio-economic policy challenges. The EU Instrument for Pre-accession Assistance (IPA) supports reforms in the enlargement countries with financial and technical help. Out of the almost â‚¬650 million destined to Albania over the programming period 2014-2022, â‚¬44 are supporting competitiveness and innovation. Policy simulations using CGE modelling techniques show positive macro-economic effects of the IPA funds for competitiveness and innovation both in the short run and in the long run thanks to productivity improvements.
Olga Diukanova
Andrea Conte
Simone Salotti
rhomolo, region, growth, impact assessment, modelling, R&D, R&I, Western Balkans, Albania, investment
2019-05
On the Solution of High-Dimensional Macro Models with Distributional Channels
http://d.repec.org/n?u=RePEc:chf:rpseri:rp1901&r=cmp
Importance of distributional channels in macroeconomic dynamics has been the object of considerable attention in empirical studies. Despite significant amount of effort aimed at incorporating heterogeneity into macroeconomics, however, their explicit inclusion in the standard policy toolbox is far from widespread. A relevant obstacle, in such cases, is the computation of equilibria. I propose a global solution method for the computation of infinite-horizon, heterogeneous agent macroeconomic models with aggregate uncertainty. Details of the algorithm are illustrated by presenting its application to a an example model: in it, aggregate dynamics depends explicitly on firm entry and exit, and individual choices are often constrained by a form of market incompleteness. Existing computational strategies are either unfeasible or provide inaccurate solutions. Moreover, global solutions are computationally expensive because the minimal representation of the aggregate state space - and thus the aggregate law of motion - faces the curse of dimensionality. The proposed strategy thus combines adaptive sparse grids with a cross-sectional density approximation, and introduces a framework for solving the more general class of dynamic models with firm or household heterogeneity accurately.
Luca MAzzone
2019-01
Machine Learning on EPEX Order Books: Insights and Forecasts
http://d.repec.org/n?u=RePEc:arx:papers:1906.06248&r=cmp
This paper employs machine learning algorithms to forecast German electricity spot market prices. The forecasts utilize in particular bid and ask order book data from the spot market but also fundamental market data like renewable infeed and expected demand. Appropriate feature extraction for the order book data is developed. Using cross-validation to optimise hyperparameters, neural networks and random forests are proposed and compared to statistical reference models. The machine learning models outperform traditional approaches.
Simon Schn\"urch
Andreas Wagner
2019-06
Market Impact and Performance of Arbitrageurs of Financial Bubbles in An Agent-Based Model
http://d.repec.org/n?u=RePEc:chf:rpseri:rp1929&r=cmp
We analyse the consequences of predicting and exploiting financial bubbles in an agent-based model, with a risky and a risk-free asset and three different trader types: fundamentalists, noise traders and "dragon riders" (DR). The DR exploit their ability to diagnose financial bubbles from the endogenous price history to determine optimal entry and exit trading times. We study the DR market impact as a function of their wealth fraction. With a proportion of up to 10%, DR are found to have a beneficial effect, reducing the volatility, value-at-risk and average bubble peak amplitudes. They thus reduce inefficiencies and stabilise the market by arbitraging the bubbles. At larger proportions, DR tend to destabilise prices, as their diagnostics of bubbles become increasingly self-referencing, leading to volatility amplification by the noise traders, which destroys the bubble characteristics that would have allowed them to predict bubbles at lower fraction of wealth. Concomitantly, bubble-based arbitrage opportunities disappear with large fraction of DR in the population of traders.
Rebecca Westphal
Didier Sornette
financial bubbles, agent-based model, arbitrageurs, prediction, noise traders, fundamen- talists, market impact
2019-06
Islanded Microgrid Operation Based on the Chaotic Crow Search Algorithm
http://d.repec.org/n?u=RePEc:pra:mprapa:94278&r=cmp
This paper investigates the optimal operation of the islanded microgrid. In order to find the optimal solution and also provide a fast response, a new heuristic method, which is known as the chaotic crow search optimization algorithm is developed. To show the merit of the model, it is tested on the IEEE 30 bus test network.
Khazaei, Ehsan
Jamaledini, Ashkan
Power system operation, Distribution Grid, Power system Economics, Energy market
2019-06-02
Incremental Risk Charge Methodology
http://d.repec.org/n?u=RePEc:pra:mprapa:94581&r=cmp
The incremental risk charge (IRC) is a new regulatory requirement from the Basel Committee in response to the recent financial crisis. Notably few models for IRC have been developed in the literature. This paper proposes a methodology consisting of two Monte Carlo simulations. The first Monte Carlo simulation simulates default, migration, and concentration in an integrated way. Combining with full re-valuation, the loss distribution at the first liquidity horizon for a subportfolio can be generated. The second Monte Carlo simulation is the random draws based on the constant level of risk assumption. It convolutes the copies of the single loss distribution to produce one year loss distribution. The aggregation of different subportfolios with different liquidity horizons is addressed. Moreover, the methodology for equity is also included, even though it is optional in IRC.
Xiao, Tim
Incremental risk charge (IRC), constant level of risk, liquidity horizon, constant loss distribution, Merton-type model, concentration.
2019-05-08
Optimal Operation of Islanded Microgrid Operation Based on the JAYA Optimization Algorithm
http://d.repec.org/n?u=RePEc:pra:mprapa:94279&r=cmp
Islanded microgrid (MG) is one of the most important challenges in the power system operation as the network can be safe and disconnected from the conjected area. Also, in the case that the market price is high, the islanded MG can have a lower operational cost by islanding from the main grid. However, optimal operation of the islanded MG is very challenging as the MG is a nonlinear problem. Hence, this paper proposed a new heuristic method known as the JAYA optimization algorithm to solve the problem. Finally, the proposed model is examined on a modified IEEE 30 bus test network to show the merit of the model.
Khazaei, Ehsan
Jamaledini, Ashkan
Power system operation, energy management, electricity market, power trading
2019-06-02
FinTechs and the Market for Financial Analysis
http://d.repec.org/n?u=RePEc:chf:rpseri:rp1910&r=cmp
Market intelligence FinTechs aggregate many data sources, including nontraditional ones, and synthesize such data using artificial intelligence to make investment recommendations. Using data from a market intelligence FinTech, we evaluate the relationship between the FinTech data coverage and market efficiency. We find an increase in price informativeness for stocks with higher FinTech coverage and that traditional sources of information have less impact on prices for those stocks. Consistent with FinTechs changing investors' behavior, we show a substitution between traditional information sources and FinTechs using internet click data. Overall, our results suggest the rise in FinTechs for investment recommendations benefits investors.
Jillian Grennan
Roni Michaely
Fintech, FinTechs (financial technology firms), Market intelligence, Artificial intelligence, Aggregators, Social media, Financial blogs, Information and market efficiency, Big data, Machine learning, Datamining, Data signal providers
2019-03
A New Nonconvex quadratic programming Technique: Practical and Fast Solver Method
http://d.repec.org/n?u=RePEc:pra:mprapa:94335&r=cmp
There exist many problems that nonconvex which are hard to solve. To overcome the nonconvexity of the problems, this paper presents a novel YALMIP-based nonconvex quadratic programming model to overcome the nonconvex problem. The proposed method is accurate, and no need to convexify the problem. Finally, some results are presented to show the effectiveness and merit of the model.
Soltani, Ali
Tashakor, Behnam
Convex optimization, math algorithm, unit commitment
2019-01
Solving the Grid-Connected Microgrid Operation by Teaching Learning Based Optimization Algorithm
http://d.repec.org/n?u=RePEc:pra:mprapa:94276&r=cmp
In this paper, the grid-connected operation of microgrid is investigated where the microgrid can exchange power with the main grid. The proposed problem is modeled as the mixed-integer linear programming (MILP) and is solved by an evolutionary algorithm known as the teaching learning-based optimization (TLBO). Finally, the proposed model is tested on a modified IEEE 33 bus test system to show the performance of the method.
Jamaledini, Ashkan
Khazaei, Ehsan
Bitaraf, Mohammad
Microgrid, TLBO, Optimization, Grid-connected operation
2019-06-02
The Keys of Predictability: A Comprehensive Study
http://d.repec.org/n?u=RePEc:chf:rpseri:rp1915&r=cmp
The problem of market predictability can be decomposed into two parts: predictive models and predictors. At first, we show how the joint employment of model selection and machine learning models can dramatically increase our capability to forecast the equity premium out-of-sample. Secondly, we introduce batteries of powerful predictors which brings the monthly S&P500 R-square to a high level of 24%. Finally, we prove how predictability is a generalized characteristic of U.S. equity markets. For each of the three parts, we consider potential and challenges posed by the new approaches in the asset pricing field.
Giovanni Barone-Adesi
Antonietta Mira
Matteo Pisati
Markets Predictability, Machine Learning, Model Selection
2019-03
Option Pricing via Multi-path Autoregressive Monte Carlo Approach
http://d.repec.org/n?u=RePEc:arx:papers:1906.06483&r=cmp
The pricing of financial derivatives, which requires massive calculations and close-to-real-time operations under many trading and arbitrage scenarios, were largely infeasible in the past. However, with the advancement of modern computing, the efficiency has substantially improved. In this work, we propose and design a multi-path option pricing approach via autoregression (AR) process and Monte Carlo Simulations (MCS). Our approach learns and incorporates the price characteristics into AR process, and re-generates the price paths for options. We apply our approach to price weekly options underlying Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) and compare the results with prior practiced models, e.g., Black-Scholes-Merton and Binomial Tree. The results show that our approach is comparable with prior practiced models.
Wei-Cheng Chen
Wei-Ho Chung
2019-06
The Sales Based Integer Program for Post-Departure Analysis in Airline Revenue Management: model and solution
http://d.repec.org/n?u=RePEc:aeg:report:2019-05&r=cmp
Airline revenue management (RM) departments pay remarkable attention to many different applications based on sales-based linear program (SBLP). SBLP is mainly used as the optimization core to solve network revenue management problems in RM decision support systems. In this study we consider a post-departure analysis, when there is no more stochasticity in the problem and we can tackle SBLP with integrality constraints on the variables (SBIP) in order to understand which should be the best possible solution. We propose a new formulation based on a market-service decomposition that allows to solve large instances of SBIP using LP-based branch-and-bound paradigm. We strengthen the bound obtained with the linear relaxations by introducing effective Chvatal-Gomory cuts. Main idea is to optimally allocate the capacity to the markets by transforming the market subproblems into a piecewise linear objective function. Major advantages are significant reduction of the problem size and the possibility of deriving a concave objective function which is strengthened dynamically. Numerical results are reported. Providing realistic integral solutions move forward the network revenue management state of the art.
Giorgio Grani
Gianmaria Leo
Laura Palagi
Mauro Piacentini
Hunkar Toyoglu
revenue management ; mixed-integer programming ; decomposition ; airline ; piecewise linear
2019
Should Faustmann forecast climate change?
http://d.repec.org/n?u=RePEc:ces:ceswps:_7636&r=cmp
Climate change is predicted to substantially alter forest growth. Optimally, forest owners should take these future changes into account when making rotation decisions today. However, the fundamental uncertainty surrounding climate change makes predicting these shifts hard. Hence, this paper asks whether forecasting them is necessary for optimal rotation decisions. While climate-change uncertainty makes it theoretically impossible to calculate expected profit losses of not forecasting, we suggest a method utilizing Monte-Carlo simulations to obtain a credible upper bound on these losses. We show that an owner following a rule of thumb - ignoring future changes and only observing changes as they come - will closely approximate optimal management. If changes are observed without too much delay, profit losses and errors in harvesting are negligible. This means that the very complex analytical problem of optimal rotation with changing growth dynamics can be simplified to a sequence of stationary problems. It also implies the argument that boundedly-rational agents may behave “as if” being fully rational has traction in forestry.
Johan Gars
Daniel Spiro
climate change, decision making under uncertainty, forestry, quantitative methods
2019