Projects

• Train a Bitcoin trading agent with reinforcement learning: because of the unbalanced scenario in Bitcoin market, we try to train an agent with reinforcement learning to find the arbitrage to make profit. The agent’s performance could make positive revenue without considering the trancsation fees.
• Tennis Result Prediction via Bayesian Models and Monte Carlo Methods: by analyzing the historical data of the two players to get the distribution of the goal in each serve. By control, the variance of such a distribution, we can predict the result through the Markov Chain Monte Carlo.
• Livestock Weight Prediction: The Forecaster Livestock Team, financially supported by Agriwebb, will improve the prediction of livestock growth based on biophysical models and machine learning.
• Foragecaster(demo): is an innovative tool that integrates machine learning algorithms with biophysical models to accurately predict livestock weights. This approach enhances the precision of weight forecasts for better livestock management.
• Foragecaster is an innovative tool that integrates machine learning algorithms with biophysical models to accurately predict livestock weights. This approach enhances the precision of weight forecasts for better livestock management.