Thank you very much for your presentation. It is a very interesting study.
I wonder if the causality analysis is possible. If your method can predict the variables that affect the final quality, I guess it seems possible to clarify the main cause that dominantly determines the quality. Is there any reason why such an approach cannot be adopted?
Thank you for your question.
As shown in the proceedings, the R^2 value of each control variable and the quality characteristic (seat hardness) is very low. From this, it is difficult to predict the control variables to affect the final quality.
Moreover, the temperature and humidity around the production equipment would be likely to affect the seat hardness. On the other hand, it is difficult to keep the temperature and humidity of the facility where the production equipment is located constant, so it is necessary to change the value of the control variables depending on the season.
Dear Dr. Tanizaki, thank you for your interesting approach of hybrid design of machine learning and optimisation. In the presentation, you mention a large number of variables (15 variables) which can influence the hardness of the automotive seat made of a polymeric foam materials. Then, you also mention the influence of temperature and humidity. Are temperature and humidity included in the 15 variables or are they extra “external” variables to take into consideration?
Dear Prof. Teti, Thank you for your question.
I apologize that the presentation is difficult to understand because the control variables are not specifically described due to the confidentiality agreement with the joint research company.
The temperature and humidity in the 15 variables.
Thank you very much for your presentation. It is a very interesting study.
I wonder if the causality analysis is possible. If your method can predict the variables that affect the final quality, I guess it seems possible to clarify the main cause that dominantly determines the quality. Is there any reason why such an approach cannot be adopted?
Thank you for your question.
As shown in the proceedings, the R^2 value of each control variable and the quality characteristic (seat hardness) is very low. From this, it is difficult to predict the control variables to affect the final quality.
Moreover, the temperature and humidity around the production equipment would be likely to affect the seat hardness. On the other hand, it is difficult to keep the temperature and humidity of the facility where the production equipment is located constant, so it is necessary to change the value of the control variables depending on the season.
Dear Dr. Tanizaki, thank you for your interesting approach of hybrid design of machine learning and optimisation. In the presentation, you mention a large number of variables (15 variables) which can influence the hardness of the automotive seat made of a polymeric foam materials. Then, you also mention the influence of temperature and humidity. Are temperature and humidity included in the 15 variables or are they extra “external” variables to take into consideration?
Dear Prof. Teti, Thank you for your question.
I apologize that the presentation is difficult to understand because the control variables are not specifically described due to the confidentiality agreement with the joint research company.
The temperature and humidity in the 15 variables.