ABSTRACT:In industrial forest plantations, the spatial distribution of management units for harvest scheduling influences the timber production cost and the nonrenewable resources consumption, due to issues related to transport logistic. In this context, this research aimed to formulate Integer Linear Programming (ILP) by means of the application of Floyd-Warshall network optimization algorithm to generate timber production routes, minimizing the production costs resulting from harvest activities and forest road maintenance. Then, scenarios were simulated considering different minimal harvest ages for Pinus spp. and Eucalyptus spp. stands. The planning horizon was five years with annual periodicity. The study area was 23,330 hectares of forests, located in Paraná state (southern Brazil). We compared the simulated scenarios according to the following parameter indicators: harvest income, building road network and the production unit cost. The decreasing of the minimal harvest age reduces the mean production of management units scheduled to be harvested, in other hand, it requires fewer roads to be built, and consequently increases the production unit cost. The solutions obtained by using ILP models presented an optimality gap lower than 0.1%. PLANO DE AGENDAMENTO DA COLHEITA FLORESTAL INTEGRADO À REDE DE ESTRADASRESUMO: Nas plantações florestais industriais, a distribuição espacial das unidades de manejo agendadas para colheita influencia o custo de produção madeireira e o consumo de recursos não renováveis, devido às questões relacionadas com a logística de transporte. Neste contexto, o objetivo do estudo foi formular modelos de Programação Linear Inteira (PLI) por meio da aplicação do algoritmo de otimização em redes Floyd-Warshall na geração de rotas de produção madeireira, minimizando o custo de produção resultante das atividades de colheita e manutenção de estradas. Em seguida, foram simulados cenários considerando distintas idades mínimas de corte de talhões de Pinus spp. e Eucalyptus spp. O horizonte de tempo foi de 5 anos com periodicidade anual. A área de estudo foi 23.330 hectares de florestas, localizados no estado do Paraná (sul do Brasil). Nos comparamos os cenários simulados de acordo com os seguintes parâmetros indicadores: rendimento da colheita, rede de estradas utilizada e custo unitário de produção. A diminuição da idade mínima de corte reduz a produtividade média das unidades de manejo agendadas para colheita e, por outro lado, requer menos estradas edificadas, o que consequentemente diminui o custo unitário de produção. As soluções obtidas pelo uso dos modelos PLI estiveram menos de 0,1% distantes do ótimo global.

No abstract

No abstract

The Transportation Problem is one of the classical problems in Operations Research. The development of more powerful personal computers in the past years made it possible to solve larger problems, creating huge opportunities of applications. The MODI method is the most cited method in the literature to solve the Transportation Problem. This method requires an initial basic feasible solution and for this purpose the literature often presents the Northwest Corner Rule, the Lest Cost Method and the Vogel Method. This paper aims to study these three different methods and analyze the number of iterations necessary to reach the optimal solution when they are used.

The increasing demand for information classification based on content in the age of social media and e-commerce has led to the need for automated product classification using their descriptions. This study aims to evaluate various techniques for this task, with a focus on descriptions written in Portuguese. A pipeline is implemented to preprocess the data, including lowercasing, accent removal, and unigram tokenization. The bag of words method is then used to convert text into numerical data, and five classification techniques are applied: argmaxtf, argmaxtfnorm, argmaxtfidf from information retrieval, and two machine learning methods logistic regression and support vector machines. The performance of each technique is evaluated using simple accuracy via thirty-fold cross validation. The results show that logistic regression achieves the highest mean accuracy among the evaluated techniques.