基于数据挖掘的天气因素研究——以吉安县为例

RESEARCH ON WEATHER FACTORS BASED ON DATA MINING–TAKING JI’AN AS AN EXAMPLE

影响天气的因素有降雨量、气压、气温、风速及风向等,研究天气影响因素对人们的生产生活有一定的指导作用。本研究提出使用数据挖掘技术探究天气因素之间的相关性。使用K-S检验方法对数据进行正态性检验,使用Pearson相关系数,对天气逐日数据降雨量、气压、气温、风速进行相关性检验,使用Eta相关系数对天气逐日数据降雨量、气压、气温、风速与风向进行相关性检验,使用一元线性回归对天气因素气压和气温进行分析。在中国气象网站的吉安县1980-2019年天气数据上进行实验,实验结果表明:数据集中各数据项不服从正态性分布,天气当中的气压与气温具有强负关联线性关系,风向类型与气压、气温有强关联关系。

This paper proposed to use data mining technology to explore the correlation among weather factors.Pearson correlation coefficient and Eta correlation coefficient tests were performed on daily data of rainfall,air pressure,air temperature,wind speed and wind direction.The linear equation was used to fit the air pressure and air temperature with high correlation.Experiments were carried out on the weather data of Ji’an city from China meteorological website.The experimental results showed that Pearson correlation coefficient analysis suggested air pressure was strongly negative correlated with air temperature,but Eta correlation coefficient analysis suggested wind direction type is strongly correlated with air pressure and air temperature.