Optimize XGBoost hyperparameters via diagnostic-driven Optuna TPE search.
npx skills add https://github.com/aliyun/qwen-dianjin --skill xgb-tuning
Please help me install this Agent Skill. Skill: xgb-tuning Source: https://github.com/aliyun/qwen-dianjin/tree/main/DianJin-SKILLS/financial-engineering-expert/xgb-tuning Command: npx skills add https://github.com/aliyun/qwen-dianjin --skill xgb-tuning
帮你在 XGBoost 基线模型基础上，针对过拟合或欠拟合进行有方向性的超参数调优，并通过诊断报告让你掌控每一轮迭代是否值得继续。
Ask the AI to tune the XGBoost hyperparameters for your binary classification dataset using the prepared baseline model, with inputs provided as --data_path, --target, and --features, and to run one diagnostic-guided tuning round in interactive mode for approval.
High-intent search queries and answers about installing and using this skill.
Optimize XGBoost hyperparameters to control binary classification overfitting by applying diagnostic-driven constrained Optuna TPE search. This workflow evaluates train and validation data gaps to dynamically tighten or expand the search space, producing a structured final parameter set that balances generalization.
Diagnostic-driven hyperparameter tuning for risk modeling is a constrained Optuna TPE search workflow that evaluates KS, AUC, and overfitting gaps per round. It outputs structured diagnostic reports to guide whether to continue iterations, ensuring the final parameters meet finance-style constraints.
Yes, you need a trained baseline model before tuning XGBoost hyperparameters. This Skill requires pre-existing train, validation, and out-of-time data to perform diagnostic-driven optimization, avoiding blind parameter search without a baseline reference.
Tune XGBoost parameters using Optuna for finance datasets by running an interactive or AUTO mode search that evaluates KS and AUC metrics. The workflow uses validation data for early stopping and out-of-time data to verify the final parameter set under finance-style constraints.
Limitations of diagnostic-driven tuning include strict boundaries against feature exploration and engineering, focusing solely on hyperparameter optimization. It requires separate train, validation, and out-of-time datasets, and depends on a pre-trained baseline model to function correctly.
ovachieverProvide standardized drug discovery datasets with train/validation/test splits.
aliyunRun XGBoost, LR+WoE, and DNN on a shared split and aggregate OOT metrics.
aliyunJointly optimize WoE binning and logistic regression regularization with Optuna TPE.
brainbytes-devQuantify uncertainty in trading parameter estimates and portfolio optimization using Bayesian methods.
