GitHub - federicakulka/learning_hierarchical_compositional_structure_tree_generated_data: This repository contains the codes relating to my master's thesis: Learning Hierarchical and Compositional Structure from Tree-Generated Data. Specifically, these codes were used: to solve the saddle-point equations; to generate hierarchical data; to train the learner and compute test performance for it and for the baselines; to plot results.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
analyze_rs_outputs.py		analyze_rs_outputs.py
bp_finite_l_vs_popdyn.py		bp_finite_l_vs_popdyn.py
bp_popdyn_vs_gaussian_census.py		bp_popdyn_vs_gaussian_census.py
broadcasting_trees_RS.jl		broadcasting_trees_RS.jl
census_vs_bp_finite_ell.py		census_vs_bp_finite_ell.py
census_vs_gaussian_approx.py		census_vs_gaussian_approx.py
common.jl		common.jl
compare_rs_vs_experiments.py		compare_rs_vs_experiments.py
experiment_ising_sklearn_lbfgs_precomputed_data_v4.py		experiment_ising_sklearn_lbfgs_precomputed_data_v4.py
generate_sequences_q2.py		generate_sequences_q2.py
population_dynamics_BP.py		population_dynamics_BP.py

About

This repository contains the codes relating to my master's thesis: Learning Hierarchical and Compositional Structure from Tree-Generated Data. Specifically, these codes were used: to solve the saddle-point equations; to generate hierarchical data; to train the learner and compute test performance for it and for the baselines; to plot results.