SECureAi

We used Isolation Tree to detect unusual patterns within the data and transcend it to the SHAP model in order to understand why did the Isolation Tree detect those anomalies. We later created visuals to depict the data and created Tables for the SEC to recognize that there is a slight potential of fraudulent insider trading. We used streamlit to represent our findings. Some of the challenges we faced was collecting the Twitter data and mapping the different APIs which share the same SEC database. We were unable to get Twitter data however we still manage to track and detect unusual patterns within a large Dataframe. The Accomplishments we are most proud of is how we paired the Isolation Tree with SHAP and the streamlit visuals we made for the SEC. We learned how to correctly collect data into a csv and reorganize unique values of that column into more columns. We also learned how to map the CIK to make our dataframe more accurate. Next thing we want to do is sentiment analysis and potentially help the SEC fight fraudulent insider trading by making a rule based system with NLP.

Built With

• python