search

Akshat111111 / Hedging-of-Financial-Derivatives

This strategy works for every market condition irrespective of the movement
BSD 3-Clause "New" or "Revised" License
72 stars 101 forks source link

💡[FEATURE]: Crypto Market Anomaly Detection #560

Closed aishikaranjan closed 3 weeks ago

aishikaranjan commented 1 month ago

Is your feature request related to a problem? Please describe. Yes, the feature is related to the problem of detecting anomalies in the cryptocurrency market. The volatile nature of cryptocurrencies makes it crucial to identify unusual patterns or changes that could indicate potential risks, fraudulent activities, or significant market events.

Describe the solution you'd like Firstly, I would perform EDA to identify potential features for anomaly detection. Then preprocess the data and implement logistic regression model to establish a baseline and subsequently evaluate the model performance using metrics such as F1-score, confusion matrix, and classification report.

Describe alternatives you've considered Multiple models can be combined. Ensemble techniques like Random Forests or Gradient Boosting can be used to handle complex relationships and improve the generalization.

github-actions[bot] commented 1 month ago

Hi there! Thanks for opening this issue. We appreciate your contribution to this open-source project. We aim to respond or assign your issue as soon as possible.