如何使用机器学习进行仿真? [英] How to do simulation using machine learning?
问题描述
大家好,
我是Azure和ML的新手,但已经观看了许多视频并亲自进行了一些实验.
i'm new to Azure and ML but have already watched many yt videos and made some experiments myself.
我有一组具有独立变量的数据,这些变量是二进制和数字分类的.我的因变量是二进制.
I have a set of data with independent variables which are binary and numeric categorical. My dependent variable is binary.
我进行了特定设置的实验(A),并收集了有关购买行为的数据.我现在更改的是实验(B)的A的设置,参与者保持不变.我的目标是使用我收集的数据来预测购买行为 实验A.
I conducted an experiment (A) with a specific set up and collected data about buying behavior. I changed the set up of A which is experiment (B) now, the participants stay the same. My goal is to predict the buying behavior using the data I collected in experiment A.
我创建了一个ML实验,以在实验A的设置下训练模型.对模型的评估得出确定系数:.82
I created an ML experiment to train a model with the setting of experiment A. The evaluation of the model resulted in a coefficient of determination: .82
我正在寻找答案:
1.如何在不运行实验的情况下模拟实验B的结果.
1. How to simulate the results for experiment B without running the experiment.
2.如何将模拟实验B的结果与实验B的真实结果进行比较,以比较结果并确定模拟是否准确.
2. How to compare the results of the simulated experiment B with the real results of experiment B in order to compare the results and to find out whether the simulation was accurate.
先谢谢您!
亨德里克
< style type ="文本/css"& lt; style type ="文本/css"</style>
<style type="text/css"><style type="text/css"></style>
推荐答案
论坛中的其他用户可能还有其他建议,但我认为创建实验副本并使用新的数据集重新运行以比较结果是最简单的.
Other users in the forum might have some additional suggestion, but I think it's simplest to create a copy of your experiment and re-run it with the new dataset to compare the results.
此致,
Jaya
Regards,
Jaya
这篇关于如何使用机器学习进行仿真?的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!