Energy Prediction Knowledge Discovery Model Based On Gradient Boosting Machines

"Energy" is the foundation of human social development.The earth's energy storage is limited.The excessive exploitation and use of energy by humans can cause serious damage to the ecological environment and social environment,such as environmental pollution,intensified contradiction between energy supply and demand.Therefore,how to adjust the structure of energy consumption and energy production and accelerate the transformation of energy is particularly important.Forecasting energy consumption is one of the important factors in adjusting energy structure,and it can even lead to the future energy structure.At present,energy prediction is mainly conducted through traditional knowledge discovery algorithms.Traditional algorithms are inherently poor in robustness and data compatibility,and are difficult to adapt to changing energy environments,making it impossible to accurately predict energy sources.Artificial intelligence(AI),as a hot topic in the current scientific community,advocates that computers can effectively simulate human thinking and achieve intelligent processing of data or things.Gradient Boosting(GB)is one of the important machine learning algorithms for artificial intelligence.Good compatibility and extensibility make it a popular algorithm for data mining.The great advantages of the GB algorithm in forecasting and classification provide new ideas for energy prediction,and it also provides technical support for the construction of accurate and reliable energy discovery knowledge discovery models.This article combines the gradient-boosting algorithms in the field of artificial intelligence and related theories and techniques of knowledge discovery to predict the trend of energy consumption in the scientific community in recent years.First of all,this paper analyzes the research results of gradient elevating algorithms and knowledge discovery at home and abroad,and analyzes the defects and deficiencies of existing research.Then,according to the general steps of traditional knowledge discovery,the knowledge discovery process of energy prediction is designed in detail.On this basis,based on the theory and technology of knowledge discovery,this paper builds a GBM-based energy prediction knowledge discovery model using Gradient Boosting Machine(GBM),an integrated gradient-boosting algorithm,which mainly includes users and data.Acquisition,data preprocessing,data mining,pattern assessment,and results visualization are six modules.The model designs the rules of feature engineering,selects,analyzes,and compares the energy-related feature variables,filters out the appropriate features,and constructs a multiple linear regression energy prediction model.At the same time,the model evaluation mechanism was designed to evaluate the prediction results in all aspects of energy prediction,and the evaluation results were fed back to each module of the model to improve the accuracy of knowledge discovery.Finally,experiments were carried out to predict the future electricity consumption trend of New York State through the analysis of historical data of energy consumption in New York State over the years.The forecast results provide a good basis for power plants and power supply companies to formulate energy strategies.Through experiments,the GBM-based energy prediction knowledge discovery model proposed in this paper can quickly and accurately predict the energy consumption trend of energy in the future,improve the efficiency and accuracy of energy prediction,and prove the feasibility and practicality of the proposed model.