This sub section is dedicated to collabs , workshops, conferences and reports done by myself during my Bachelor and Master degrees.

• Principal Component Analysis (Portuguese)
It is intended through this multivariate technique to reduce the variables to the smallest possible number (dimensionality reduction), by a new set of orthogonal variables (main components) and not correlated with each other, in relation to the nutritional information of pizzas (the dataset in analysis). Some criterias are explained, the retention of the number of componentes, nuances in the weights of the eigenvector interpretations and the different rotation methods are exposed in this article.
• Cluster Analysis (Portuguese)
The study under review refers to lifelong learning/education in different member states of the European Union. It is intended to detect if the different periods of membership of the respective countries has some impact on how economic agents' incomes are applied in constant learning. The method used is the Hierarchical since we have no suspicion of how many K clusters exists, which would be necessary in the non-hierarchical method and because we are faced with a small dataset.
• Multilevel Models (Portuguese)
An exposition of the advantages and disadvantages of using Linear Hierarchical Models or multilevel models are presented. In what contexts and frameworks should they be used and when to prefer them over other models. The data contained in Hox (2018) is used as an illustration.
• Monte Carlo Simulation in Linear Regression (Portuguese)
We run a Monte Carlos simulation in order to estimate the regressors of the Simple Linear Regression Model. A short exposition is done in order to check how we can improve regressors for the model in hand and generate a better fit.
• Visualizing statistical models: Removing the blindfold a critic. (Portuguese)
The papar is a critic to the methods and processes used by Wickham, Cook and Hofmann in their approach of visualizing statistical models.