课程: AutoML: Build Production-Ready Models Quickly!
今天就开通帐号，24,100 门业界名师课程任您挑！
Improving your AutoGluon model - Python教程
课程: AutoML: Build Production-Ready Models Quickly!
Improving your AutoGluon model
- Once we know how to build and evaluate models in AutoGluon, we can turn our attention to customizing and improving them. In this video, we will explore some of the actions that can help us meet business objectives. I have imported our data set and split it as usual. I've also kept our initial configurations for evaluation metric and label, while updating the save part, so we don't override our previous models. We have shown that we can prototype and evaluate a lot of models quickly using AutoGluon. However, it is important to note that AutoML is no substitute for domain knowledge of the problem at hand. It is important to bring that knowledge to whatever problem we are building AutoML models for. One area we can apply domain knowledge, is Feature Engineering. To exemplify this, we've created a new feature, age of building, from our data. We have also cast year of observation…
随堂练习，边学边练
下载课堂讲义。学练结合，紧跟进度，轻松巩固知识。
下载课程离线学习
下载课程到移动设备，轻松离线学习。在领英学习 iOS 或安卓版 APP 下载课程