Posts
-
SIR Modeling: The Math Behind an Epidemic Curve, and How to Simulate One
Mathematics turns out to be a useful tool for the quantitative analysis of infection dynamics, the same way it’s useful for predicting any other observable physical phenomenon. Future infection rates can be predicted with real accuracy, provided we’re working from accurate current conditions, and that should offer some comfort: it means we can watch how our behavior today shapes the pandemic’s future, and how changes in that behavior push the curve one way or the other.
-
Eight Puzzle: State Space Search and the Cost of Heuristic Choices
Everyone’s familiar with the classic 8 Puzzle game:
-
Jane Street Puzzle: March 2026: Planetary Parade
I’m posting this a little late but I want to work through some of the older puzzles as practice before I move on to the current ones. Here’s the problem as stated on the Jane Street website:
-
CPI Forecasting: Recurrent Neural Network Methods
I want to answer the question of whether or not you can use a recurrent neural network to predict CPI. The headline series everyone cares about is CPIAUCSL, the seasonally adjusted Consumer Price Index for All Urban Consumers. More specifically: can a sequence model extract more signal than a naive autoregression, and what assumptions you have to make about the data along the way for the answer to mean anything?
-
Forecasting Implied Volatility: A Hybrid ARIMA-XGBoost Approach
One of the more interesting problems in options trading is forecasting how implied volatility (IV) will evolve over the remaining life of a contract. If you can do this, you can identify mispricings in the options market and trade against them systematically.
subscribe via RSS