Skip to content

Aterg264/TelegramBot

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

interDiaBetGPT

Project inicialized in J ON: The Beach Hackaton with Intersystems.

The idea of this project is to develop a Telegram bot utilizing ChatGPT, which is capable of answering complex questions about statistical data related to healthcare. To do it, in our first version of the project, we integrated Iris IntegratedML and Large Lenguage Model ChatGPT.

During the model validation process we have seen that the variance, R2, etc. metrics were not good. We have created other models by varying the number of columns used and reducing the diabetes values to (0 --> the patient does not have diabetes, 1--> the patient has diabetes). We finally got the 'DiabetesModelValidate31' model, which had the best probability metric statistics. Finally we have created a persistent field in the 'ValidateDataDiabetes' table that contains the result of the prediction for each row. This last table is the table on which we perform the SQL queries to answer the questions made by the user.

To build the Docker image: docker build -t hackathon-app .

To run the container: docker run -it --rm --name hackathon-app hackathon-app

Data Source: https://www.kaggle.com/datasets/cdc/behavioral-risk-factor-surveillance-system

Details of each feature: https://www.cdc.gov/brfss/annual_data/2015/pdf/codebook15_llcp.pdf

Frameworks: https://docs.langchain.com/docs/

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 95.4%
  • Dockerfile 4.6%