Census and Inventory Method of Pollution Sources Based on Big Data Technology under Machine Learning

In order to strengthen the supervision and management of environmental pollution timely understand record basic information potential enterprises institutions, this paper proposes a general survey inventory method sources based on big data technology under machine learning. Firstly, evaluates screens provided by government departments, constructs learning classification model, uses variety algorithms compare analyze according idea deal with practical problems. Then, calibration set constructed is used as training predict classify national industrial commercial data, provincial municipal departments. The experimental results show that naive Bayesian algorithm best algorithm, F1 values each are increased 32.92%, 21.42%, 14.91%, respectively. classified prediction screened Internet shows accuracy final supplementary 17.26%, which similar city. availability model established in proven.