Research On Price Prediction Of Ceramic Artworks Based On Data Mining

Since the reform and opening up,the Chinese art market has developed rapidly,with ceramic artworks being an important component.Due to the dual material and cultural qualities of ceramic artworks,more and more people are paying attention to ceramic artworks.In addition,ceramic artworks not only meet people’s aesthetic enjoyment,but also bring economic benefits.However,due to the media’s priceless attitude towards "artworks" the frequent hype surrounding the price of ceramic art,such as "artists are cash cows" has led to a mysterious veil over the price of ceramic art,causing the general public to have little knowledge of the price of ceramic art,thereby restricting the development of the ceramic art market.Therefore,this article proposes three price measurement models for ceramic artworks,allowing everyone to have a better understanding of the price of ceramic artworks and contributing to the healthy development of the ceramic art market.Firstly,the decision tree prediction model was used to quantify the ceramic features using the assignment method.The prices of the top 160 ceramic artworks in Appendix 1 were predicted,and the results showed that the predicted fit of the model was 0.85.Among them,about 80% of the predicted prices of ceramic artworks had errors within 0.1,indicating that the model can accurately predict the prices of ceramic artworks.The second is the random forest prediction model.The average method was used to quantify the ceramic characteristics,and the price of 100 ceramic artworks in the160-260 data in Appendix 1 was predicted.The predicted fitting degree reached 0.98,of which about 93% of the ceramic artworks’ predicted price errors were within 0.1,or even 65% of the errors were within 0.05.The prediction accuracy was high.The third is the multi-layer perceptron prediction model.The standard deviation method is used to quantify the ceramic characteristics and predict the prices of the last100 ceramic artworks in Appendix 1.The results show that the fitting degree of the model is 0.93,of which about 90% of the predicted price errors of ceramic artworks are within 0.1,and the prediction results are relatively ideal.Finally,the decision tree,random forest and multi-layer perceptron models are used to predict and compare the same 50 data.The prediction results show that the prediction error of 41 data in the random forest model is smaller than that of the decision tree;The prediction error of 32 data in the multi-layer perceptron model is smaller than that of the decision tree,and the average prediction error of each model is 0.0673,0.045,and 0.051 respectively.It can be seen that the prediction accuracy of random forest and multi-layer perceptron in the three models is higher than that of the decision tree,and the prediction accuracy of random forest is the highest.In summary,the three models mainly studied the price prediction of ceramic artworks from the perspectives of data mining and data analysis,and all of them can relatively accurately predict the price of ceramic artworks.Therefore,the model in this article can not only provide guidance for ceramic art trading,but also provide theoretical support for the ceramic market,thereby promoting the healthy development of the art market.