Data Mining for Seasonal Influences in Broiler Breeding Based on Observational Study

摘要

For the modern poultry breeding companies, it is worthwhile to extract valuable knowledge from the massive historical data to help future production and management. However, data analysis and mining of poultry raising dataset is a challenge due to the complexity and uncertainty bring by the influence of environmental and physiological factors. In this paper, data mining based on observational study is proposed for the research of seasonal influences in broiler breeding. Systematic observational study with the statistical analysis and data mining technology is adopted including macro analysis, exploratory data analysis, and modeling and prediction. Case study using the broiler growth dataset of the most famous poultry raising company in China shows the effectiveness of our approach.