In the traditional sense, when a fire breaks out, it takes a lot of time for a fire to go out to put out the fire after receiving the news. Due to the lack of timely response, it is difficult to control the fire from time to time. In this model, the possibility is minimized by the statewide networking system composed of SSA UAV and repeater. This model has the advantages of rapid response, low cost, high accuracy and sensitivity. We use entropy weight method and machine learning to find the most influential factor on aircraft price and proportion. Instead of using a single algorithm, a series of algorithms is applied, and there is a conjoint principle that each property of them is independent from the value of any other property. Naive Bayes consider that there is no relationship between the property and each of the property independently make contribution to the probability. Notwithstanding, there is a drawback of naïve Bayes algorism that properties are not independent with each other invariably. That is to say, we can forecast a class by using probability which provides sets of properties because of Naive Bayes algorithm. Naive Bayes algorithm requires less training compared with the other classification methods. The only work that should be done before predicting is to find the parameters of individual probability distribution of the property, which can be done fast and explicitly. This implies that even for high-dimensional data points or large amounts data points, naive Bayes classifier can perform well. Grasp the main factors, ignore the secondary factors, simplify the model to find a suitable ratio.

Yunjie Qu, Yichen Song. Using entropy weight method and machine learning to improve the allocation of rescue resources in case of fire. International Journal of New Developments in Engineering and Society (2021) Vol.5, Issue 2: 8-12. https://doi.org/10.25236/IJNDES.2021.050203.

[1] Laboratory of Statistical and Mathematical Methodology Division of Computer Research and Technology, National Institutes of Health, Bethesda, MD 20205, USA: 112-119.

[2] Lta G. G. Kreft, Jan de Leeuw. Introducing multilevel modeling .SAGE, Jun 18, 1 998

[3] T. A. B. Snijders, Roel J. Bosker .Multilevel analysis: an introduction to basic and ad- vanced multilevel modeling. SAGE, Dec 7, 1999

[4] Ronald H. Heck Scott Loring Thomas .An introduction to multilevel modeling tech- niques.L awrence Erlbaum Associates, 2000

[5] DavidKaplan.Structuralequationmodeling: foundationsandextensions. SAGE. 2009