基于24种脂质标志物降解动力学的指纹遗留时间预测模型研究OA

Predicting the age of fingerprint impressions can provide a basis for inferring the time of crime occurrence and evaluating the reliability of testimonies,effectively narrowing down the scope of investigation.In this study,the content ranges of 23 kinds of fatty acids and squalene in fingerprint impressions were quantitatively analyzed using gas chromatography-mass spectrometry(GC-MS).By combining 24 types of lipids in pairs,276 lipid ratio feature sets were constructed.Spearman's univariate analysis was used to screen ratio features significantly correlating with the age of fingerprint impressions.The Gini importance scores of the random forest algorithm were combined to explore the nonlinear contribution of features to the prediction model,ultimately constructing a fingerprint age prediction model based on lipid ratio features.The results showed that the model could effectively distinguish fingerprint impressions left for 0,9,and 13 d,and exhibited good cross-individual robustness.The overall model performance had a coefficient of determination(R2)of 0.785 and a root mean squared error(RMSE)of 2.10 d.This study provided new technical insights and quantitative evidence for accurate prediction of fingerprint age.