Visibility Prediction Based On XGBoost And Markov Chain Combined Model
Hui Tang, ZhenHan Wei, JiangLi Liu · 2021 7th International Conference on Computer and Communications (ICCC) · 2021
Improving the accuracy of visibility prediction is conducive to taking corresponding measures for weather with different visibility to achieve the purpose of ensuring the safety of life and property. Meteorological factors, environmental factors, terrain and human activities all have an impact on visibility. At present, visibility prediction mainly includes meteorological map analysis, numerical model forecasting, statistical forecasting and other methods. This paper proposes a visibility prediction based on Extreme Gradient Boosting and Markov chain combined model (MC-XGBoost). Firstly, build the MC-XGBoost model, the visibility related feature parameters are selected as the input of the model, and the corresponding visibility is taken as the output of the model. Then perform model training, adjust the model parameters to make the model achieve the optimal prediction effect. Finally, the trained model is applied to visibility prediction. Experimental results show that MC-XGBoost model can effectively predict visibility with high accuracy. This model can be used to predict visibility, which provides a certain reference value for the construction of a visibility prediction model in the future.