We analyze expenditure patterns of discretionary funds by Brazilian congress members. This analysis is based on a large dataset containing over $7$ million expenses made publicly available by the Brazilian government. This dataset has, up to now, remained widely untouched by machine learning methods. Our main contributions are two-fold: (i) we provide a novel dataset benchmark for machine learning-based efforts for government transparency to the broader research community, and (ii) introduce a neural network-based approach for analyzing and visualizing outlying expense patterns. Our hope is that the approach presented here can inspire new machine learning methodologies for government transparency applicable to other developing nations.

Being an analyst working on statistical analysis, more often than none, there is almost always a need to "explain" how did you come up with the conclusion with the techniques applied? The most demanding part of all this is: " you must explain this like I am 5 years old." Oki, so let me try to explain one machine learning algorithm called "randomforest" for a classification problem that is recognised as a black-box algorithm! Let's get started with a metaphor that hopefully is as close to everyday life as possible... In a congress, say that there are 100 members in total, they need to vote to decide whether they are going to pass a new law or not.