Machine learning for financial transaction classification across companies using character-level word embeddings of text fields

An important initial step in accounting is mapping financial transfers to the corresponding accounts. We devised machine-learning-based systems that automate this process. They use word embeddings with character-level features to process transaction texts. When considering 473 companies independently, our approach achieved an average top-1 accuracy of 80.50%, outperforming baselines that exclude the transaction texts or rely on a lexical bag-of-words text representation. We extended the approach to generalizes across companies and even across different corporate sectors. After standardization of the account structures and careful feature engineering, a single classifier trained on 44 companies from 28 sectors achieved a test accuracy of more than 80%. When trained on 43 companies and tested on the remaining one, the system achieved an average performance of 64.62%. This rate increased to nearly 70% when considering only the largest sector.     

Predicting Household Poverty: A Methodological Note with a Kenyan Example

Household surveys provide data that is used for identifying and measuring the poverty status of households and individuals. However, carrying out such surveys is expensive, especially in poor developing countries. Thus it is important to make maximum use of the available survey data in developing countries, especially in sub‐Saharan Africa, where such data are expensive to collect and analyse. This paper develops a simple method for using poverty indices derived from survey data for a given year, to predict poverty rates for subsequent periods without having to conduct a new household survey. We illustrate the workings of the method with data from Kenyan household surveys for 1994 and 1997.     

The Local Whittle Estimator of Long-Memory Stochastic Volatility

We propose a new semiparametric estimator of the degree of persistencein volatility for long memory stochastic volatility (LMSV) models.The estimator uses the periodogram of the log squared returnsin a local Whittle criterion which explicitly accounts for thenoise term in the LMSV model. Finite-sample and asymptotic standarderrors for the estimator are provided. An extensive simulationstudy reveals that the local Whittle estimator is much lessbiased and that the finite-sample standard errors yield moreaccurate confidence intervals than the widely-used GPH estimator.The estimator is also found to be robust against possible leverageeffects. In an empirical analysis of the daily Deutsche Mark/USDollar exchange rate, the new estimator indicates stronger persistencein volatility than the GPH estimator, provided that a largenumber of frequencies is used.     

Information styles of marketing and R&D personnel during technological product innovation projects

Abstract The research question in the present article can be phrased in the following way: what are the elements which influence the perception of the utility of information received from another function? A field study research has been conducted in 40 Belgian companies. In each company, we have studied two on-going innovation projects (one planning project, one development project). Crossfunctional communication behaviours at the R&D/marketing interface have been measured by means of mailed structured questionnaires. Three-hundred and eighty-six questionnaires have been returned. The data analyses show that there are four underlying information dimensions, i.e. the perceived relevance, the perceived comprehensibility, the perceived novelty and the perceived credibility of information. We also discuss some contingency variables (function of the message receiver, stage in the innovation process) that might moderate the impact of these dimensions. It will be shown that the interaction between source and receiver has a significant impact on the perceived credibility, the perceived comprehensibility and the perceived novelty of the received information.     

Optimal immigration, education and growth in the long run

The paper extends Manning's model on education and balanced growth to include labour immigration. Each immigration unit is assumed to consist of one skilled worker and some unskilled members. The optimal immigration policy which maximizes the per capita steady-state consumption of the host country is derived. We show that optimal immigration policy can reduce the steady-state skilled labour ratio. More interesting still, contrary to the widespread belief that immigration of skilled workers hurts local skilled workers, it is the unskilled local workers whose interests are threatened by optimal immigration policy.     

Rural traffic management: The Burrator Reservoir experiment

The introduction of traffic management schemes has been one response of the National Parks to the problem of ever-increasing car-borne visitors and the consequent damage caused to the rural environment. By reviewing these schemes and by determining and explaining attitudes towards one particular traffic management experiment, the paper aims to deduce lessons for future policy. The analysis, based on two visitor surveys, suggests that the likelihood of success is improved when schemes are integrated, containing both ‘carrot’ and ‘stick’ elements. Since public perceptions are also found to be pivotal to its success, appropriate marketing of a scheme is vital. The paper concludes, however, that to achieve the fundamental objective of greater accessibility with less mobility, a more wide-ranging marketing effort is required intended to engender a sea change in attitude towards both the car and public transport.     

The approximation capabilities of the minflex-Laurents,translog, and generalized Leontief forms

This paper compares the approximation capabilities of the minflex-Laurents translog and minflex generalized Leontief cost functions with their translog and generalized Leontief counterparts in Monte Carlo experiments. The minflex Laurent specifications generally provided closer approximations to underlying technical and economic parameters. Imposition of nonlinear restrictions on some of the parameters of the minflex Laurent models yielded measurable improvement in estimated elasticities of substitutions, returns to scale, and rates of technical change.The refereeing process of this paper was handled through E. Appelbaum.     

Price Changes and Some Underlying Aspects of Measurement of Poverty

The paper deals with effects of price changes on measurement of poverty. It is shown that there are situations, when prices and nominal incomes change, a higher nutritional intake or an increase in commodity consumption of the poor can be associated with higher poverty gaps and a higher poverty index of the society. A positive relationship of this kind is quite counter-intuitive in the context of poverty measurement. This is caused by the adjustment of the poverty line and/or incomes of one situation by the prices of the other for the purpose of comparison.