This repository started with the goal of documenting the progress of the Financial WebApp we are developing to fulfill the requirements of Data Science Bachelor of Arts. All work is unique

https://share.streamlit.io/bonorinoa/datascience_capping/main/WebApp/ai_investment_tool.py

After observing promising results this repository has since become the place where the app lives. We intend to continue developing with the goal of offering a high quality product that could empower retail investors.

Latest version:

• 6.0

Fixes coming up in next version (7.0):

• Web development moved away from Streamlit to proper tools. JAvascript is intended to be used, which will allow us to build more complicated models with Tensorflow.js
• The use of Twint API will be deprecated due to its unreliability. Twitter's official API v2.0 will be considered.
• Sentiment Analysis with roBERTa fine-tuned for analyzing tweets -> cardiffnlp/twitter-roberta-base-sentiment
• XGBoost and fine-tuned RNN will be added to the available Machine Learning models.
• A more comprehensive walk through the app. This has two goals: enhance FinML to serve as an educational platform, and minimize the probability of the user passing the wrong input.
• Smarter formatters. For instance, the current version only supports the exact ticker of the company in capital letters. We will add flexibility by accounting for grammatical errors and misformatted input.