Statistics: What may teleworkers and clams have in common?

Evaluation of anxiety, depression and sleep quality on teleworkers - Miguel Fonseca

Telework's massive worldwide expansion and its potential impact on mental health are crucial issues in today’s society and, in particular, in Portugal. Thus, an enquiry was taken on a group of teleworkers to evaluate anxiety, depression and sleep quality on teleworkers and explore associations between these variables, quality of life and perceived productivity. Results were analyzed using descriptive statistical methods and regression analysis. Results show that teleworkers reported anxiety and depressive symptoms with predominance of anxiety and very high levels of sleep impairment. Better sleep quality was associated with longer sleep duration and better job.

Application of regularization methods to high-dimensional data as tool for predicting the geographic origin of the saltwater clam Ruditapes philippinarum - Clara Yokochi

As a result of the growing globalization of seafood markets, traceability has become essential to safeguard food safety. With consumers being increasingly aware of the potential hazards that contaminated seafood may cause to their health, ensuring that the geographic origin of seafood is not mislabeled is a first step to fight fraudulent practices that aim to cover-up illegal fishing.

Ruditapes philippinarum is a species of saltwater clam that is commercially harvested for consumption, being one of the most important bivalve grown in aquaculture worldwide. This species' location of origin can be predicted by modeling features like their biochemical and geochemical fingerprints. The exploited dataset constitutes 30 clam samples, detailing information on 44 composition features, with the purpose of identifying which features distinguish between three geographic origins: Ria de Vigo, Ria de Aveiro, Estuário do Tejo, i.e, a classical Multinomial Logistic Regression problem. However, given the high-dimensionality of the dataset (number of variables higher than number of observations), the estimation of the model coefficients is compromised as Fisher's Information Matrix is no longer invertible. To overcome this problem, three dimensionality reduction methods were applied to model the origin of the clams: Ridge, LASSO and Elastic Net. Additionally, since datasets of only 30 samples challenge the process of model validation, the re-sampling method of Monte Carlo Cross-Validation was also implemented. We finalize comparing the results between the three methods, identifying which has the best predictive performance and comparing the estimation errors in each category of the response.

Presentation of the Statistics field as a part of the Profile in Actuarial Sciences, Statistics and Operation Research of the MSc in Mathematics and Applications - Isabel Gomes


Miguel Fonseca

Miguel Fonseca is a professor of the Mathematics Department in the NOVA School of Science and Technology, part of the NOVA University Lisbon. He has a doctor degree in Mathematics, with specialization in Statistics. His main scientific areas are experimental analysis, biostatistics and data science. He is also interested in machine learning and financial mathematics, and is an avid consumer of science fiction and music.

Clara Yokochi

Clara Yokochi has BSc in Pure Mathematics (NOVA School of Science and Technology, Portugal) and is now finishing a MSc in Applied Mathematics (NOVA School of Science and Technology, Portugal), specialized in Actuarial Sciences, Statistics and Operations Research. Her love for mathematics began the moment she realized that most problems can be, to some extent, thought of as being puzzle games. Being a very creative person, analyzing data, finding patterns, and trying to look at problems from many different perspectives to find interesting solutions has been one of her main enjoyments.