Evaluate standard machine classifiers on particle reaction data. Under construction.
Posts a data set of particle reactions labelled as "possible" or "impossible". Runs standard classifiers on this set and compares to the null space method (see paper Simultaneous Discovery of Conservation Laws and Hidden Particles With Smith Matrix Decomposition. The hypothesis is that on this dataset, domain knowledge beats algorithm.