In this project we developed a bank stock advisor. Our code runs a ML model to predict bank stock prices which then allows us to give/receive advice based on a persons' portfolio. Also attached is a chatbot powered by GPT-3.5 which has been fine-tuned to answer finacial questions as an educational finacial advisor. (Note, this should not be used to trade real assests, always check with an actual certified finacial advisor before making decisions that can affect your finances)
- Operating Systems: Mac OS, Windows
- Programming Language: Python
- Libraries: Pandas,Numpy,Hvplot,Matplotlib,Warnings,Watermark,Panel,Standard Scaler,Streamlit
- Frameworks: JupyterLab,HTML
Access our application online without the need for local installation. Simply click on the following link:
This version is identical to the one you can set up locally following the instructions provided below.
If you wish to install our code, there are a couple of things that need to initally be installed which include.
After all of the things above have been installed you can then proceed with cloning the code(to either a pre-created folder or a new folder).
- Go the the github repository and locate the green code button
- Click on the code button and then copy the HTTPS link provided
- Open a terminal and navigate to where you would like the code to be i.e.
cd .\project3\. - Run the
git clone https://github.com/greg-krulin/Project-2-.gitcommand
project2reco.mp4
Once the code has been copied and your pre-requisites have been installed then you need to download your module to properly run code. Found below you can see a list of modules and under the list you can see how to install each module.
List of modules: Modules
-1. Activate a conda dev environment within your terminal
conda activate dev
-2. Install the modules within your terminal
pip install -r requirements.txt
- Launch Streamlit (make sure your web browser is not in dark mode)
-Open up a terminal on your desktop
-Navigate into the folder in which you cloned the repository
-Search for the ml_app.py file and when located use thestreamlit run ml_app.pycommand
Untitled.video.-.Made.with.Clipchamp.mp4
- Explore features within the streamlit
-Change your selections within the stock graphs, correlation, multi-correlation and machine learning results tabs to recieve different results
explore.mp4
- Chat with the bot and receive bank stock data/advice
-Locate the miniatrue purple message bubble and begin asking/answering questions
ChatBot.mp4
-
Greg Krulin - Lead code developer and researcher
Email: grgr279@gmail.com -
Mark Beers - Lead research
Email: beers.mark@gmail.com -
Chris Cummock - Research and advisor
Email: ccummock@gmail.com -
Eyasu Alemu - Streamlit and chatbot production
Email: bekaqa01@gmail.com -
Samuel Jew - Streamlit and chatbot production
Email: samjew95@gmail.com -
John Garcia - Github manager and researcher
Email: jdganna222@gmail.com