This notebook was part of a project in CEFET/RJ where I had the opportunity to contribute and learn with my professor Douglas Cardoso data from Brazilian researcher's profile and train a classifier to predict their profile based of patents they contributed. This work was done during July 2020 and May 2022. Important note: This notebook has sensitive data so in order to keep anonimous I hide the output. Initially I was designated to extract data from a datsheet in which has data from the degree of the researcher, and other data. I had to clean eventual incosistencies such as error of date of birth. I also plotted some histograms and scatter plot to have a graphical view of the data. Later on, there was another datasheet in which it has different information where the researchers had to answer questions and based on these answers we tried to predict the in which class the researcher fits in. In this point it was really challenging because we have to merge both of these datasheet and unpreddicted output happen, it which we had to clean and organize manually. In the end we finally trained a decision tree classifier to preddict researcher's class comparing all different classification metrics.
halliday-onice/Brazilian-researcher-s-profile-
This notebook was part of a project in CEFET/RJ where I had the opportunity to contribute and learn with my professor Douglas Cardoso data from Brazilian researcher's profile and train a classifier to predict their profile based of patents they contributed. This work was done during July 2020 and May 2022
Jupyter Notebook