Clebson C. A. Sá

I'm currently studying at the Federal University of Minas Gerais (UFMG) as a Ph.D. student in Computer Science in the database laboratory LBD, which allowed me to grasp the fundamentals of machine learning, data science, statistics and a strong basis in Project and Analysis of Algorithms.

As it's possible to ponder, my research interests include applied machine learning, statistics and data science. Currently, I'm pushing forward the findings of my master's dissertation to optimize the predictive performance of ensembles of boosted additive bagged trees such as Bagging, Random Forests and Boosting for information retrieval.

Things I Can Do

I'm quite versitile!
I have worked with most of the state of the art classification and regression algorithms such as Support Vector Machines, Decision Trees, Bagging, Random Forests, and Boosting to name a few.
As a consequence of that, I acquired some super powers as a Software Engineer, Developer and Data Scientist. Some of those skills are:

• Excelled at coding in: C/C++, Java,
  Python, and Bash
• Deep understanding of Quantitative Methods in Experimental Analysis
• Read books and scientific papers with ease, enjoying and having a lot of fun in the way.
• Strong Love for C/C++ while solving problems in Hacker Rank
• Power to have fun while working with hard problems.
• Power to express, hear and work in group.

A Few Accomplishments

So far, I was able to publish a few scientific papers. A brief summary of the three last published papers follows bellow:

Generalized BROOF-L2R: A General Framework for Learning to Rank Based on Boosting and Random Forests

In this work we combine Random Forests with Boosting in order to leverage the predictive performance in the Learning to Rank (L2R) realm. By combining these two algorithms, we were able to devise 4 loss functions for the regression scenario, in which we were able to surprisingly surpass all baselines with statistical significance gains.

Learning to Rank with Boosting and Random Forests: An Hybrid Model

In this paper, we derive an adaptation of the BROOF algorithm for the task of L2R with an extensive benchmark comparison of several state state-of-the-art algorithms. The experiments shows that our devised algorithm using regression surpassed most of the baselines with statistical significance.

Educative Software to assist the Teaching/Learning process of Math in APAEs.

In this paper we showed the benefits of having a Web Software to assist students with mental illness at APAEs in the community of Rio Paranaíba. The Web Software was composed of a series of ludic activities pre configured according to the level of each student by the teacher, influencing the students to think about what was seen in class in the computer lab without the dependency of a specialized professional in the computer area.

Prizes and awards

Learning Online

I manage to increase my abilities by learning lots of stuff online. Some of the coolest things that I learned online is on Coursera. How great can MOOCs be? I'm not sure, but I loved all of those courses:

Do you Love Code?

I have some pieces of code to share with you. Don't be shy, take a look at them. By the way, who doesn't love git code snippets? Check these repositories: