Bhils

Project repository for brainhack in Rome 2025. Bhils stands for "Brain Hacking In Latent Space".

For the brain hackathon event, we analysed both a toy dataset and a MRI dataset.

The toy dataset is made of two independent variables X (height $X_1$ and distance to work $X_2$), two dependent variables (weight $Y_1$ and shoe size $Y_2$) and nine samples.

The toy dataset X and Y variables are related through different linear correlation coefficients values: $r_{Y1X1}=0.89$, $p-value=0.0012$; $r_{Y1X2}=0.06$, $p-value=0.87$; $r_{Y2X1}=0.91$, $p-value=0.0006$; $r_{Y2X2}=0.05$, $p-value=0.89$; $r_{X1X2}=0.07$, $p-value=0.84$; $r_{Y1Y2}=0.78$, $p-value=0.01$.