https://community.jmp.com/t5/Discussions/Seek-the-probability-of-achieving-the-maximum-and-minimum-pass/m-p/906670#M106504 <P>Hi Wenwen,</P> <P>&nbsp;</P> <P>This is a binary response case. EDA and modeling can be applied to seek optimal combination for A and B.</P> <P>Attached is the jmp file. You can run the scripts to see the results.</P> <P>Since this is an imbalanced dataset with few "Fail" records, modeling with Logistic Regression, Generalized Regression and Random Forest in JMP PRO are not able to achieve good performance for predicting Y now. So I really recommend you combine the results of EDA and the models to narrow your process window. Also, I think confirmation experiments are needed to verify the parameters.</P> <P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="分析数据 - Graph Builder.jpg" style="width: 780px;"><img src="https://community.jmp.com/t5/image/serverpage/image-id/84672i27EF47A832F4FF62/image-size/large?v=v2&amp;px=999" role="button" title="分析数据 - Graph Builder.jpg" alt="分析数据 - Graph Builder.jpg" /></span></P> <P>&nbsp;</P> <P>Best,</P> <P>Joyce Pan</P> Thu, 09 Oct 2025 08:40:49 GMT Joyce_Pan 2025-10-09T08:40:49Z https://community.jmp.com/t5/Discussions/Seek-the-probability-of-achieving-the-maximum-and-minimum-pass/m-p/906638#M106501 <P>I have a set of data with already determined results. I want to use the known determination conclusions to infer what combinations of AB data_before can achieve the highest and lowest probabilities of passing the test.</P> Thu, 09 Oct 2025 07:02:44 GMT https://community.jmp.com/t5/Discussions/Seek-the-probability-of-achieving-the-maximum-and-minimum-pass/m-p/906638#M106501 wenwenzhou 2025-10-09T07:02:44Z https://community.jmp.com/t5/Discussions/Seek-the-probability-of-achieving-the-maximum-and-minimum-pass/m-p/906670#M106504 <P>Hi Wenwen,</P> <P>&nbsp;</P> <P>This is a binary response case. EDA and modeling can be applied to seek optimal combination for A and B.</P> <P>Attached is the jmp file. You can run the scripts to see the results.</P> <P>Since this is an imbalanced dataset with few "Fail" records, modeling with Logistic Regression, Generalized Regression and Random Forest in JMP PRO are not able to achieve good performance for predicting Y now. So I really recommend you combine the results of EDA and the models to narrow your process window. Also, I think confirmation experiments are needed to verify the parameters.</P> <P><span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="分析数据 - Graph Builder.jpg" style="width: 780px;"><img src="https://community.jmp.com/t5/image/serverpage/image-id/84672i27EF47A832F4FF62/image-size/large?v=v2&amp;px=999" role="button" title="分析数据 - Graph Builder.jpg" alt="分析数据 - Graph Builder.jpg" /></span></P> <P>&nbsp;</P> <P>Best,</P> <P>Joyce Pan</P> Thu, 09 Oct 2025 08:40:49 GMT https://community.jmp.com/t5/Discussions/Seek-the-probability-of-achieving-the-maximum-and-minimum-pass/m-p/906670#M106504 Joyce_Pan 2025-10-09T08:40:49Z