Coursera Spring Sale
40% Off Coursera Plus Annual!
Grab it
Explore symbolic learning and rule extraction through a hands-on tutorial featuring the Sole.jl Julia package in this 25-minute conference talk from JuliaCon Local Paris 2025. Learn how symbolic learning differs from neural networks by creating interpretable models that can be translated into logical rules, making them more readable and explainable than traditional statistical approaches. Discover the comprehensive SOLE framework (SymbOlic LEarning), an open-source Julia ecosystem that guides users through the entire process from initial data preprocessing to symbolic model creation and rule extraction. Master the integration of Sole.jl with popular packages including DecisionTree.jl, ModalDecisionTrees.jl, MLJ.jl, and XGBoost.jl for building decision trees and ensemble models like random forests. Understand how to work with non-tabular data such as images and time-series by leveraging SoleData.jl to interpret datasets as logical interpretations using modal logic frameworks. Get introduced to two new additions to the SOLE ecosystem: ModalAssociationRules.jl for mining association rules between instances, and SolePostHoc.jl for extracting, interpreting, and simplifying rule sets from symbolic models. Follow practical demonstrations showing how to fit decision tree models, extract logical rules, and manipulate them for enhanced interpretability, emphasizing the framework's user-friendly design and comprehensive capabilities for symbolic machine learning applications.
