Use of Multiple Data Sources in Collaborative Data Mining

Agriculture is one of the domains that depend on weather forecasts and which would improve its performance if some features could be predicted. Because of that, we try to define a concept that is capable of generating predictive results for temperature, based on multiple sources and use of virtual machine learning. We define collaborative approaches in data mining that use multiple sources and we analyze the results from a comparative point of view with the independent process. For each approach, we use a machine learning algorithm which is applied to the combination of data sources from the weather stations. The research proposes a model of the data mining process for a collaborative variant with multiple sources. This leads to the conclusion that a collaborative approach generates better results than a standalone one.