We create a ML model to predict the results of the national standardized tests Saber 11 based on the socioeconomic characteristics of the students. Our main objective was to measure the extend on which the socioeconomic status of a student determines its performance on standardized tests, a question heavily related to the meritocracy fallacy.
We study the under reporting of Covid cases in Colombia and perform some calculations to estimate the true dimensions of the Pandemic. We improved the previous reported algorithm by taking into account specifics of demographics in Colombia and the relations between age groups and Covid19 sensitivity.
We analyse students comments on classes reviews and use NLP methods to identify latent topics in the comments, also we design a data processing pipeline to automatically identify bad reviews and the topics of those reviews.
Work done for my Bachelor thesis in physics. We developed high performance code to simulate multi-lane systems with cars interactions and performed statistical analysis on two lane systems. You can take a look at the thesis in this link.
The results of this work were presented at the Traffic and Granular Flow at Navarra in 2019.