The FINAL grades
Schedule: Tuesday and Thursday: from 8am to 10am
Room: 52 IC3
This course will cover in breadth many algorithms of machine learning/data mining. We will cover, in general terms, of the following machine learning "tasks": data transformation, classification, regression and clustering.
We will cover a variety of algorithms, providing some intuitions on how, why and when they work, but for only a few of them there will be a in depth study of the mathematical formulation and properties. We believe that this breadth approach will allow the student to solve the more common problems they may face in data mining.
The course will be based on the book Inteligencia Artificial: uma abordagem de aprendizado de maquina (in Portuguese) by Faceli, Lorena, Gama & Carvalho. There will be also some extra material that complements some of the topics covered in the book. .
The course will be evaluated in two forms, to be selected by the student.
The exercises must be completed using either R (with a set of packages for machine learning) or Python (with the scikit learn package and maybe others) - details below
There will be two reports of 5 to 10 pages, to be turned in at the mid semester and at the end. The first (due date to be announced) will cover the bibliographic review of the problem you are attacking, as well as some results of the use of the "standard" machine learning algorithms on that dataset.
A final report by the end of the semester, with more detailed comparisons, with the general aim of producing a publishable paper on that problem.
R:
Python:
Discussion groups on ML: