A carregar...
Publicação
O impacto da analítica hoje e no futuro
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 858.63 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
With the exponential technological growth that has taken place over the last few decades, it is
becoming easier and easier to capture and store business data. In fact, it is practically impossible today
to find a medium or large company that does not use advanced information systems in order to improve
the management of their business. However, this significant increase in the ease of obtaining and storing
large volumes of data has not been adequately matched by the ability to interpret these same datasets.
Eventhough data stored is well structured and business-oriented, in most cases it is not easy to interpet
with conventional anlytics. It is therefore common to have cases with a large volume of data, with valid
information, which would in the future result in greater commercial value, but it is not clear how to
extract and generate useful information for business activities. In these situations, implicit patterns may
often be hiden in data history which, in turn, may contain information that is very relevant to the busi ness. It is in this context that the concept of Knowledge Discovery in Databases (KDD) emerged in 1989.
In this sense, with this work it is intended, as a global objective, to understand the role of analytics
in the business world, to learn the different applications for large sets of data. Finally, an example of
analytics application is shown to study hotel business data and to improve decision-making related to
hotel bookings.
Therefore, the aim is to create a model capable of predicting the cancellations of reservations in
advance for two hotels. To achieve this, there is a set of skills, technologies and practices that provide
insights to support management decision-making and the resolution of business challenges. As such, we
intend to create a booking cancellation prediction model, through the use of logistic regression models.
Through a set of independent variables, this methodology allows client’s characteristics and booking
reservation to predict client’s cancellations. Furthermore, it is possible to find out which variable has the
greatest information gain in the model. For the model construction, the response variable is defined as
the cancellation, or not, of the reservation, depending on the variables that best characterize its profile.
For this example we are dealing with a classification model. In this way, two distinct samples,
training and testing, are established. Thus, the set of observations of the test sample will never influence
the construction of the model and, in turn, it is possible to test and evaluate its discriminatory capacity.
Descrição
Trabalho de Projeto de Mestrado, Matemática Aplicada à Economia e Gestão, 2022, Universidade de Lisboa, Faculdade de Ciências
Palavras-chave
Analítica Regressao Logística Cancelamento de reservas de hotel Teses de mestrado - 2022