I am a scientist at Harvard Medical School. For an academic research paper , we've built two random forest models in Python. These models take in 30+ datapoints about the health status of a mouse and return predictions about its future health status. We would like to make these models available for widespread research use.
Currently the models have been saved with pickle ([login to view URL]). We would like to build a website where an user can upload an excel or csv file with each column representing a parameter in the model, and each row representing an animal.
This would ideally be done with a Django or Flask app. We attempted to make a Shiny app ourselves, using reticulate to get Python to run in R, but were unsuccessful. Likely there is a version issue. We made the models using sklearn version 19 and Python 3.7. There likely needs to be some version control issues and we would need to confirm that the models run successfully.
We also envision a tab on the web app with background information in the form of text, links, and downloadable pdf files.
We have our own server from which we can host the web app once complete.
23 pekerja bebas membida secara purata $594 untuk pekerjaan ini
Greetings, It will be an honor for me if I get a chance to work with a Harvard scientist. We can help you host your modals on google cloud for everyone to use. Regards, Hirak