تحديد اثر استخدام المبيدات على التلوث البيئي بالطرق الطيفية و الكروماتوغرافية جنوب شرق الجزائر أنموذجا

Abstract

Despite the important role played by pesticides in combating pests in all their forms and afflicting agricultural crops and their contribution to raising production quantitatively and qualitatively and achieving food security in the light of the world's high population density، excessive and indiscriminate use of them heralds a catastrophe affecting various ecosystems. (Water، dirt،air)، where the name pesticide was associated with the presence of malignant and chronic diseases. Many studies have demonstrated the toxic impact of pesticides on humans and vital enemies. In this regard، the study aims to identify the reality of the use and circulation of pesticides and to measure farmers' awareness of the risks of pesticides and how to deal with them and the precautions to be taken during their application to crops in four states in the south-eastern region of Algeria. (Biskara،elouaed، and ouargla،elmenia)، which is experiencing prosperity and an important agricultural revolution، This was through the introduction of a questionnaire covering five axes، analyzed by the statistical analysis program SPSS، where the alpha-chrombach coefficient، the Pearson correlation coefficient and the modular and intermediate computational deviation were calculated. The fact that pesticides are toxic is an inevitable result as toxic chemicals that vary in toxicity from one pesticide to another. Many new pesticides are put on the market in different types and forms. Their toxicity must be studied and determined. The latter is a challenge due to the substantial resources required to evaluate a chemical in the living body.This study drew on a hybrid model combining a new model called the Long Term Memory Model (LSTM) and the linear regression model (MLR) to predict the acute toxicity of oral GN LD50 in mice for chemical pesticides based on the contribution of collective interaction (GIC) developed using a database of 303 pesticides belonging to different chemical groups. The dataset was randomly divided 80  training kit and 20  a test kit of a total database where the modules developed using the best selected parameters achieved interesting results and high accuracy in predicting the toxicity of LD50 in the test stages with the R2 binding coefficient equal to 0.8888، 0.1312294، 0.362256، 0.2926325، 2045653، 3.99981،MSE،RMSE،MAE،RAE،RRSE