Feature engineering: The missing link in applying machine learning to deliver business value

Hendra Suryanto shares a case study from a Canadian financial lender that his company helped transition from manual to automated credit decisioning, using gradient boosting machine and deep learning to build the model. In addition to modeling techniques, Hendra highlights the role feature engineering plays in improving model performance.

本讲话将用英语授课，同时会提供中文同声传译。中文版本摘要会在英文摘要下面给出。 Hendra Suryanto shares a case study from a Canadian financial lender that his company helped transition from manual to automated credit decisioning, using gradient boosting machine and deep learning to build the model. The process begins with prototyping, moves to production and automation, and ends up at operationalization, which involves translating predictions into decisions by incorporating the business rules and handing them over to the operations and business teams. In addition to modeling techniques, Hendra highlights the role feature engineering plays in improving model performance. 这是一个来自于加拿大的一家金融借贷公司的案例研究。我们帮助他们完成了从手工贷款审批到自动化的转变。整个过程从原型化开始，然后是生产部署后的自动化，再结合业务规则把预测转变成审批决策，并最终交付给运营和业务团队使用。 我们应用梯度增强机和深度学习来够建模型。除了建模技术，我们会特别强调特征工程在改进模型表现里所起的重要作用。 我们在行业专家的指导下手工进行了特征工程，这展示了创建好的特征是如何对交付更好的产出产生影响的。我们也探索了使用深度学习和增强学习来辅助特征工程。