ML.NET回归

本文介绍了ML.NET回归 - 预测的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！ 问题描述

您好，我正在审核和测试ML.NET机器学习项目中的示例项目"GITHUB"

到目前为止，它似乎已在一定程度上得到了很好的记录。如果我按照这些例子，我将能够调整各种类型的预测数据，但在示例（出租车费用）中，您似乎可以根据许多变量预测票价。

我想知道如何预测"何时"只使用1到几个参数就可能发生一些事情。例如。

假设我有碰撞数据。

Datetime，IntersectionId，CollisionSpeed，Injuries，死亡，IntersectionType，CollisionRating

有了这个，我可以修改TaxtFare ML示例，但我需要大部分数据来预测"CollisionRating"

如果我想知道"何时"怎么办？可能会发生下一级别10碰撞评级或任何其他数据类型？

或者即使可能发生NEXT碰撞，也没有输入。

可能??

$

解决方案

对于ML上的问题，您需要在他们的GitHub页面上提出问题：

github/dotnet/machinelearning-samples

github/ dotnet / machinelearning-samples / issues / 86

问候， Jaya

Hello, I am reviewing and testing the sample projects in the ML.NET machine learning project on "GITHUB"

So far it seems well documented to an extent. If I follow the examples I would be able to adjust the data for various types of predictions but in the examples (Taxi Fares) it seems you can predict the fare based on a number of variables.

I'd like to know how to predict "when" something would potentially happen using only 1 to few parameters. Example.

Let's say I have collision data.

Datetime, IntersectionId, CollisionSpeed, Injuries, Deaths, IntersectionType, CollisionRating

With this,I could modify the TaxtFare ML example but I would need most of the data to get a prediction on "CollisionRating"

What if I wanted to know "when" the next level 10 Collision Rating would possibly happen or any of the other data types?

Or even when the NEXT collision would potentially happen, no input.

Possible??

解决方案

Hi,

For questions on ML you would need to raise the issue on their GitHub page:

github/dotnet/machinelearning-samples

github/dotnet/machinelearning-samples/issues/86

Regards, Jaya