Advisoring

PhD Pedro Henrique da Silva Pinto

Interior Point Methods and quasi-Newton strategies.

PhD César Postingel Ramos

Kriging models and applications to Chemical Engineering problems.

MsC Mariana Maronezzi Brezovsky

Optimization and SVD.

UNDERGRAD Giovana Melo dos Santos - Water and transportation problems using IPMs

In this work, we model water distribution problem and some transportation problems using (mixed integer) linear programming, branch and price and interior point techniques.

MsC Pedro Henrique da Silva Pinto - A quasi-Newton predictor-corrector IPM

In this work, based on quasi-Newton ideas implemented in IPMs, we proposed a predictor-corrector primal-dual interior point method which applies Broyden "Bad" updates as a local strategy to improve centrality. We compared our implementation in Julia against the Tulip.jl software.

UNDERGRAD Giovana Melo dos Santos - The hunger games problem

In this work, we studied modeling techniques to solve the "Hunger games problem", a kind of knapsack problem with constraints. Three approaches were developed: a MILP formulation using JuMP, dynamic programming and a graph-based one. See code in GitHub.

PhD Anderson Schwertner - Derivative-free trust-region Low-Order Value Optimization

UNDERGRAD Mariana Maronezzi Brezovsky - SVD factorization and recommendation systems

UNDERGRAD Julia Guizardi - Efficient update of matrix factorizations

In this work we, studied and implemented efficient techniques for updating the LU and QR factorizations, under some specific changes of the original matrix.

UNDERGRAD Joaquim Martins - Strategies for the irregular packing problem

In this work, we will model the packing of irregular items by means of nonlinear programming models for the triangle packing problem.

UNDERGRAD Pedro Henrique Pinto and Joaquim Martins - Projected gradient methods and applications

UNDERGRAD Joaquim Martins - Irregular packing problems

Continuous models for irregular packing problems into rectangular containers.

MsC Lucas Moschen - Application of Newton's and Broyden's methods to linear programming problems

UNDERGRAD Joaquim Martins - Secant method applied to packing problems

In this project we will study the theoretical properties of the secant method for solving nonlinear equations in one variable. The method will be applied to simplified versions of several packing problems, in particular those for which the derivatives are difficult to obtain. Numerical implementations in Python will be provided.

UNDERGRAD Julia Guizardi- Multivariable Newton polynomial and applications to the design of experiments

UNDERGRAD Pedro Henrique Pinto - Packing of polygons using nonlinear programming

In this work we will apply nonlinear programming techniques to pack polygons inside polygonal areas, maybe allowing rotations.

MASTER Edilaine - LOVO (Lower Order-value Optimization) Levenberg-Marquardt algorithms

UNDERGRAD Matheus Souza - Derivative-free algorithm for a solar tracker

In this work we will study and develop the direct search derivative-free optimization algorithm for embedded systems. The algorithm will run on an Arduino board in order to maximize the total amount of solar energy obtained in solar panels that are able to follow the sun.

UNDERGRAD Eloiza - Development of a low cost solar tracker using optimization algorithms

In this

UNDERGRAD Julia - Interpolation techniques to the optimization of experiments

UNDERGRAD Pedro - Packing of polygons using linear programming

In this work we will extend the previous results in square packing to the packing of convex polygons inside convex polygonal regions. The goal is to maximize the number of packed items.

UNDERGRAD Pedro - Square packing using linear programming

Modeled the problem of packing squares into convex polygonal regions using linear programming. Using this idea a simple heuristic was developed to pack several squares into convex polygonal regions, without rotating the square. A GeoGebra simple program was developed, so users can model any convex region as a polygonal one and pack squares on it. Generator, code and tests

UNDERGRAD Matheus - Newton method for dummies

Related the Babylonian Method for finding square roots of natural numbers and the Newton method for roots of polynomials. The goal was to make the Newton method available to the general audience. Produced 2 interactive programs: the first one using Scratch and the second one through a GeoGebra notebook. Download codes.