A new package would provide a text input describing a system, event, or scenario, and return a structured analysis based on Nassim Taleb's concepts of Black Swan events and antifragility. It would ide

In the presence of black swan events, an auto-encoder trained on normal data will fail to reconstruct them accurately, resulting in abnormally high residual errors. These errors can be modeled by POT to flag potential anomalous events or black swans.

Official PyTorch implementation of the paper “Catching Both Gray and Black Swans: Open-set Supervised Anomaly Detection”, open-set anomaly detection, few-shot anomaly detection, semi-supervised anomaly detection.

Collaborative Project with group members using LSTM (Long-Term/Short-Term Deep Learning) models to predict stock prices.

Expected sigma event frequency of S&P 500 daily returns based on historical returns, compared to normal distribution expected frequency.