RESUMO

Sugarcane mills in Brazil collect a vast amount of data relating to production on an annual basis. The analysis of this type of database is complex, especially when factors relating to varieties, climate, detailed management techniques, and edaphic conditions are taken into account. The aim of this paper was to perform a decision tree analysis of a detailed database from a production unit and to evaluate the actionable patterns found in terms of their usefulness for increasing production. The decision tree revealed interpretable patterns relating to sugarcane yield (R2 = 0.617), certain of which were actionable and had been previously studied and reported in the literature. Based on two actionable patterns relating to soil chemistry, intervention which will increase production by almost 2 % were suitable for recommendation. The method was successful in reproducing the knowledge of experts of the factors which influence sugarcane yield, and the decision trees can support the decision-making process in the context of production and the formulation of hypotheses for specific experiments.

RESUMO

Sugarcane mills in Brazil collect a vast amount of data relating to production on an annual basis. The analysis of this type of database is complex, especially when factors relating to varieties, climate, detailed management techniques, and edaphic conditions are taken into account. The aim of this paper was to perform a decision tree analysis of a detailed database from a production unit and to evaluate the actionable patterns found in terms of their usefulness for increasing production. The decision tree revealed interpretable patterns relating to sugarcane yield (R2 = 0.617), certain of which were actionable and had been previously studied and reported in the literature. Based on two actionable patterns relating to soil chemistry, intervention which will increase production by almost 2 % were suitable for recommendation. The method was successful in reproducing the knowledge of experts of the factors which influence sugarcane yield, and the decision trees can support the decision-making process in the context of production and the formulation of hypotheses for specific experiments.(AU)

RESUMO

The present research hypothesized that the thermal, lighting and acoustic environments in commercial swine farrowing rooms vary over time and from crate to crate. This study was conducted on 27 replicates in two commercial farrowing rooms in North Central Indiana, each equipped with 60 farrowing crates. Temperature, relative humidity, light intensity, sound intensity, and air velocity were continuously monitored and estimated for each crate at the sow level, for 48 h post-farrowing, which is usually a critical period for piglet survivability. Average daily temperature for all the crates in Room 1 was 24.1 ± 2.0 °C, 1.0 °C lower (p 0.05) than in Room 2. Although the overall mean temperature was similar between rooms and seasons, frequency distribution diagrams revealed that the proportion of time spent within distinct limits of mean daily temperature ranged from 15.0 °C to 28.0 °C and varied substantially between rooms and seasons. Similar results were found for all variables measured in this study. Differences in temperature, relative humidity, light intensity, air velocity, and sound intensity in crates were as high as 9.6 °C, 57 %, 3,847.3 Lx, 0.87 m s1, and 38.7 dBC, respectively, in the same farrowing room when measured at the same instant. The results of the present research indicate that aspects that go beyond the physical environment of the sows, such as thermal, lighting, and acoustic environment can vary substantially over time and between crates of automatically climate controlled farrowing rooms. These differences should be taken into consideration in production setting and research.(AU)

Assuntos

RESUMO

The present research hypothesized that the thermal, lighting and acoustic environments in commercial swine farrowing rooms vary over time and from crate to crate. This study was conducted on 27 replicates in two commercial farrowing rooms in North Central Indiana, each equipped with 60 farrowing crates. Temperature, relative humidity, light intensity, sound intensity, and air velocity were continuously monitored and estimated for each crate at the sow level, for 48 h post-farrowing, which is usually a critical period for piglet survivability. Average daily temperature for all the crates in Room 1 was 24.1 ± 2.0 °C, 1.0 °C lower (p 0.05) than in Room 2. Although the overall mean temperature was similar between rooms and seasons, frequency distribution diagrams revealed that the proportion of time spent within distinct limits of mean daily temperature ranged from 15.0 °C to 28.0 °C and varied substantially between rooms and seasons. Similar results were found for all variables measured in this study. Differences in temperature, relative humidity, light intensity, air velocity, and sound intensity in crates were as high as 9.6 °C, 57 %, 3,847.3 Lx, 0.87 m s1, and 38.7 dBC, respectively, in the same farrowing room when measured at the same instant. The results of the present research indicate that aspects that go beyond the physical environment of the sows, such as thermal, lighting, and acoustic environment can vary substantially over time and between crates of automatically climate controlled farrowing rooms. These differences should be taken into consideration in production setting and research.

Assuntos

RESUMO

Incubatório de ovos é um setor de grande importância na Avicultura de postura. Com a redução dos custos dos equipamentos de informática cresce o armazenamento de dados para gerenciamento do processo produtivo. A Mineração de Dados surge como uma técnica para identificar conhecimentos novos e úteis nos bancos de dados. Objetivou-se, neste trabalho, explorar a técnica Arvore de Decisão em banco de dados de incubatórios de matrizes de postura, visando a elaboração de padrões de incubação. Foram disponibilizados, pela empresa Hy-Line do Brasil Ltda, dados de incubação entre os anos de 2002 e 2006 da linhagem Hy-Line W-36. Dois experimentos foram realizados. Em um deles, valores acima dos estabelecidos pela empresa como desejado para o índice "fêmeas nascidas vendáveis" foram identificados como relevantes para a geração das regras. No outro, valores abaixo dos estabelecidos pela empresa foram identificados como relevantes para a geração das regras. Foi utilizado o algoritmo Entropia C 4.5 e o software SAS-Enterprise Miner como ferramenta de análise . Como conclusão deste estudo, foi possível observar que com a técnica estudada, os dados utilizados no gerenciamento de produção são suficientes para identificar conhecimentos novos, úteis e aplicáveis a fim de melhorar a produtividade das empresas incubadoras, atendendo à demanda com diminuição do desperdício.

Hatchery is a very important sector in egg production. As computers become cheaper, there is an increase in data storage for the production management process. Data Mining has appeared as a technique to identify new and useful knowledge in databases. The objective of this work was to explore the Decision Tree technique in hatchery databases to identify the best standards of the incubation process. The data set used in this research was supplied by Hy-Line do Brasil Ltda., corresponding to the incubation period of 2002-2006, from the strain Hy-line W-36. Two experiments were carried out. In the first experiment, values higher than the company's standards for saleable females were identified as relevant to generate the rules. In the second experiment, values below those established by the company were identified as relevant for the generation of rules. Entropy C 4.5 algorithm and the software SAS-Enterprise Miner were used for data analysis. The conclusion is that, with the technique studied, the data used for production management are sufficient to identify new, useful and applicable knowledge in order to increase productivity of hatcheries, catering for the demand with less waste.

RESUMO

Um dos desafios da agricultura de precisão é oferecer subsídios para a definição de unidades de manejo para posteriores intervenções. Portanto, o objetivo deste trabalho foi avaliar os atributos químicos do solo e a produtividade da cultura de cana-de-açúcar por meio da geoestatística e mineração de dados pela indução da árvore de decisão. A produtividade da cana-de-açúcar foi mapeada em uma área de aproximadamente 23ha, utilizando-se o critério de célula, por meio de um monitor de produtividade que permitiu a elaboração de um mapa digital que representa a superfície de produção para a área em estudo. Para determinar os atributos de um Argissolo Vermelho-Amarelo, foram coletadas as amostras no início da safra 2006/2007, utilizando-se uma grade regular de 50 x 50m, nas profundidades de 0,0-0,2m e 0,2-0,4m. Os dados dos atributos do solo e da produtividade foram analisados por meio da técnica de goestatística e classificados em três níveis de produção para indução de árvore de decisão. A árvore de decisão foi induzida no programa SAS Enterprise Miner, sendo utilizado algoritmo baseado na redução de entropia. As variáveis altitude e potássio apresentaram os maiores valores de correlação com a produtividade de cana-de-açúcar. A indução de árvores de decisão permitiu verificar que a altitude é a variável com maior potencial para interpretar os mapas de produtividade de cana-de-açúcar, auxiliando na agricultura de precisão e mostrando-se uma ferramenta adequada para o estudo de definição de zonas de manejo em área cultivada com essa cultura.

One of the challenges of precision agriculture is to offer subsidies for the definition of management units for posterior interventions. Therefore, the objective of this work was to evaluate soil chemical attributes and sugarcane yield with the use of geostatistics and data mining by decision tree induction. Sugarcane yield was mapped in a 23ha field, applying the cell criterion, by using a yield monitor that allowed the elaboration of a digital map representing the surface of production of the studied area. To determine the soil attributes, soil samples were collected at the beginning of the harvest in 2006/2007 using a regular grid of 50 x 50m, in the depths of 0.0-0.2m and 0.2-0.4m. Soil attributes and sugarcane yield data were analyzed by using geostatistics techniques and were classified into three yield levels for the elaboration of the decision tree. The decision tree was induced in the software SAS Enterprise Miner, using an algorithm based on entropy reduction. Altitude and potassium presented the highest values of correlation with sugarcane yield. The induction of decision trees showed that the altitude is the variable with the greatest potential to interpret the sugarcane yield maps, then assisting in precision agriculture and, revealing an adjusted tool for the study of management definition zones in area cropped with sugarcane.