At　present,　the　problem　of　chronic　diseases　in　China　is　becoming　more　and　more　prominent,　and　there　is　a　trend　of　rejuvenation,　among　which　diabetes,　as　a　common　chronic　disease,　is　located　in　the　first　place　in　the　world　in　terms　of　the　number　of　patients.　To　alleviate　the　aforementioned　problems,　this　paper　designs　a　cloud　platform　for　intelligent　diagnosis　of　chronic　diseases.　The　platform　is　dedicated　to　helping　people　effectively　prevent　chronic　diseases　and　reduce　the　risk　of　developing　complications.　In　this　study,　diabetes　prediction　models　are　analyzed　from　multiple　perspectives,　firstly,　the　category　imbalance　problem　is　solved　by　SMOTE-ENN,　and　then　models　such　as　KNN,　logistic　regression,　random　forest,　SVM,　GBDT,　and　XGBoost　are　used　to　classify　the　samples　respectively　and　the　optimization　performances　of　the　two　hyper-parameter　tuning　methods,　Bayesian　and　random　search,　are　compared.　Experiments　show　that　XGBoost　based　on　Bayesian　optimization　outperforms　other　models　and　can　be　applied　to　intelligent　diagnosis　of　diabetes.　©　2024　IEEE.