| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 724.56 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O setor dos transportes é um setor crucial e bastante importante para o nosso dia-a-dia,
na medida em que é o único que permite transportar mercadorias e pessoas entre os
diversos países. É, portanto, um setor insubstituível e indispensável sendo uma parte
essencial do tecido funcional da nossa sociedade.
O insucesso empresarial, por sua vez, é um fenómeno que apresenta bastante interesse
para qualquer ramo, incluindo o ramo dos transportes. Tal facto tem levado a que sejam
desenvolvidos modelos de previsão do fracasso empresarial que, a partir de de um
conjunto de indicadores, indicam a probabilidade de a organização falhar.
Aliada a esta situação, assistimos diariamente a uma permanente evolução tecnológica,
bem como, a uma constante busca por métodos mais eficazes na na previsão do fracasso
empresarial, levando os investigadores a recorrer, numa tendência crescente, à
inteligência artificial.
O presente trabalho visa efetuar um estudo comparativo entre os modelos de fracasso
empresarial estatísticos e os modelos baseados em inteligência artificial para o setor dos
transportes, num horizonte temporal de 2014 a 2021 e para os países de Portugal,
Espanha, França e Itália, de modo a verificar qual dos modelos é mais eficiente. A amostra
utilizada foi constituída por um conjunto final de 4866 empresas, das quais 2881 são não
fracassadas e 1985 fracassadas.
Os modelos desenvolvidos permitiram classificar corretamente entre 71% a 73% das
empresas. Não obstante, quando estes resultados são comparados com os resultados
obtidos nos modelos de previsão do fracasso empresarial estatísticos , constata-se que
estes últimos apresentam uma capacidade preditiva superior, registando menos erros na
classificação das empresas em análise.
The transport sector is crucial and very important to our daily lives, as it is the only sector that allows us to transport goods and people between different countries. It is therefore an irreplaceable and indispensable sector and an essential part of the functional fabric of our society. Business failure, on the other hand, is a phenomenon that is of great interest to any industry, including transport. This has led to the development of business failure prediction models which, based on a set of indicators, indicate the likelihood of the organization failing. Allied to this situation, we are witnessing constant technological evolution on a daily basis, as well as a constant search for new, more effective methods that are capable of predicting business failure, leading to a growing trend towards the implementation of artificial intelligence to estimate models for predicting business failure. The aim of this paper is to carry out a comparative study between statistical business failure models and models based on artificial intelligence for the transport sector, over a time horizon of 2014 to 2021 and for the countries of Portugal, Spain, France and Italy, in order to see which of the models is more efficient. The sample used consisted of a final set of 4866 companies, of which 2881 were non-failed and 1985 failed. The results of this study made it possible to correctly classify between 71% and 73% of the companies, as well as to verify that when statistical business failure prediction models are compared with prediction models based on artificial intelligence, the latter have a higher predictive capacity, registering fewer errors in the classification of the companies under analysis as failed or non-failed.
The transport sector is crucial and very important to our daily lives, as it is the only sector that allows us to transport goods and people between different countries. It is therefore an irreplaceable and indispensable sector and an essential part of the functional fabric of our society. Business failure, on the other hand, is a phenomenon that is of great interest to any industry, including transport. This has led to the development of business failure prediction models which, based on a set of indicators, indicate the likelihood of the organization failing. Allied to this situation, we are witnessing constant technological evolution on a daily basis, as well as a constant search for new, more effective methods that are capable of predicting business failure, leading to a growing trend towards the implementation of artificial intelligence to estimate models for predicting business failure. The aim of this paper is to carry out a comparative study between statistical business failure models and models based on artificial intelligence for the transport sector, over a time horizon of 2014 to 2021 and for the countries of Portugal, Spain, France and Italy, in order to see which of the models is more efficient. The sample used consisted of a final set of 4866 companies, of which 2881 were non-failed and 1985 failed. The results of this study made it possible to correctly classify between 71% and 73% of the companies, as well as to verify that when statistical business failure prediction models are compared with prediction models based on artificial intelligence, the latter have a higher predictive capacity, registering fewer errors in the classification of the companies under analysis as failed or non-failed.
Description
Keywords
Inteligência artificial Fracasso empresarial Modelos de previsão Indicadores financeiros Setor transportes Artificial intelligence Transportation sector Business failure Financial indicator Forecasting modelss