Work / Fig 21 · Lab
Stock prediction models
A forecasting notebook using linear regression, a decision-tree classifier, and Bayesian ridge on historical prices. A study in methods, not a trading desk.
LABAI
A stock ticker under a glass dome prints prices onto a paper tape that feeds a gantry plotter. As the tape advances the pen plots the price line point by point, then draws the least-squares line through the same points.
Classical ML on financial time series
Historical prices in, predicted trends out. The stack is classical on purpose: linear regression, decision trees, Bayesian ridge. Also explored transformer and agentic workflow experiments around the same problem.
This is research code, not financial advice, and not a claim that the market was beaten.
Stack
PythonLinear regressionDecision treesBayesian ridge