ABSTRACT Air traffic management has become increasingly complex due to the increasing use of air transport. One of the main management bottlenecks is planning the efficient use of runways for takeoff and landing. This paper aims to investigate the Aircraft Landing Problem, which seeks to minimize earliness and tardiness in aircraft landing time, assigning them to a runway to land and sequencing them. A new mathematical formulation based on Job Shop was proposed for the problem, comparing it with four mathematical formulations in the literature; three directly comparable and another containing a particularity that does not allow a direct comparison with the other formulations. Computational tests were performed on 49 instances of the literature using the Gurobi Optimizer optimization package. These mathematical formulations commonly used for the ALP present difficulties in finding the optimal solution when the number of aircraft to land is large, i.e., more than 50 aircraft. Therefore, we proposed a matheuristic to solve instances with a greater number of aircraft than the Gurobi Optimizer cannot solve optimally. This matheuristic first finds an initial solution using relax-and-fix (RF) and then fix-and-optimize (FO) improves the found solution. Comparisons were also made using the first feasible solution obtained by Gurobi and then was improved with FO. Among the matheuristic variations, the one that obtained the best result was the combination of RF with FO and this also showed efficiency in relation to the work in the literature that uses FO. transport Problem time problem 4 package large ie i e i.e. 5 Therefore optimally relaxandfix relax fix (RF fixandoptimize optimize (FO variations i.e

RESUMO Estamos usando tecnologias baseadas em Inteligência Artificial em um número crescente de sistemas e ferramentas. A Inteligência Artificial pode tornar reduzir a necessidade da presença humana em muitas atividades perigosas, monótonas e cansativas, nos liberando para atividades menos perigosas e mais desafiadoras e estimulantes. Ao mesmo tempo, a Inteligência Artificial pode aumentar riscos existentes e trazer novos riscos. Para evitar ou reduzir esses riscos, é necessário o desenvolvimento de novos algoritmos de Inteligência Artificial, ou seu uso de maneiras novas e inovadoras, levando em consideração questões éticas, sociais e legais.

ABSTRACT We are using Artificial Intelligence-based technologies in an increasing number of systems and tools. Artificial Intelligence can reduce the need for human presence in many dangerous, monotonous and tiring activities, freeing us for less dangerous and more challenging and stimulating activities. At the same time, Artificial Intelligence can increase existing risks and introduce new risks. To avoid or reduce these risks, new Artificial Intelligence algorithms must be developed or used in new and innovative ways, taking into account ethical, social and legal issues.

Due to the imprecise nature of biological experiments, biological data is often characterized by the presence of redundant and noisy data. This may be due to errors that occurred during data collection, such as contaminations in laboratorial samples. It is the case of gene expression data, where the equipments and tools currently used frequently produce noisy biological data. Machine Learning algorithms have been successfully used in gene expression data analysis. Although many Machine Learning algorithms can deal with noise, detecting and removing noisy instances from the training data set can help the induction of the target hypothesis. This paper evaluates the use of distance-based pre-processing techniques for noise detection in gene expression data classification problems. This evaluation analyzes the effectiveness of the techniques investigated in removing noisy data, measured by the accuracy obtained by different Machine Learning classifiers over the pre-processed data.