nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Bayesian Linear Regression

TakumiSaruhashi

What inspired this Wish List Request?

I recently had a meeting with NIPPON SHOKUBAI CO LTD. The customer said they wanted to perform a linear regression that reflected prior knowledge, such as Bayesian linear regression. They said that linear regression often estimates coefficients that are not chemically possible.

Please describe the new or improved feature or solution you would like to see.

What would the value be to you and/or other users if this new feature is provided?

Users can perform a linear regression that reflected prior knowledge.

Is this an external customer request?

Yes. NIPPON SHOKUBAI CO LTD requested this feature.

Is there a competitive software package that has this feature or solution that we should be aware of?

Python.

https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.BayesianRidge.html

Will this impact a renewal, expansion, or adoption at a customer, and does this put an account at risk if we do not address this?

There is potential for expansion.

Status: Needs Info

Comments

Sarah-Sylvestre

SamGardner

@TakumiSaruhashi it would be helpful if we could have more information. In particular, could the customer provide a data example and to described how they would apply a bayesian approach to regression.

The link to the python package was for a method to use bayesian estimation of the regularization parameter in Ridge Regression, but the request above seems to be more general than for just using ridge regression.

TakumiSaruhashi

@SamGardner -san, Sorry for the late reply.

I checked with the customer and received the following response.

We cannot provide specific sample data.

An example of use is to build a model that is robust to outliers (a model that is assumed to be chemically valid) by incorporating chemical background (e.g., temperature has a positive effect on internal pressure) as a prior distribution on the coefficients.

We are sorry that we cannot provide specific sample data, but we can create toy data for you.

If you would like toy data, could you let us know the deadline?

SamGardner

If an example data set and description of the problem can be sent, that would be very helpful.

TakumiSaruhashi

@SamGardner -san, I will ask the customer to create the toy data.

I will contact you upon receipt of the toy data.