Interpretable Machine Learning models are essential for increasing the adoption of Machine Learning within society. One of these is called decision rules, and it provides a set of rules to describe the decision criterion used by the model. CORELS is concerned with providing the optimal set of rules given a model. In this paper CORELS will be used in order to provide rules that distinguish between two tumor types based on the genetic mutations of the patients. In the current context therefore, the decision rules model will be used to observe relationships between mutations somatic and tumor type
I modelli d’Interpretable Machine Learning sono essenziali per aumentare l’adozione del Machine Learning all'interno della società. Uno di questi viene chiamato decision rules, e fornisce una serie di regole atte a descrivere il criterio decisionale utilizzato dal modello. CORELS si occupa di fornire l’insieme di regole ottimo dato un modello. In questo documento verrà utilizzato CORELS in modo da fornire delle regole che distinguano due tipologie di tumore in base alle mutazioni genetiche dei pazienti. Nel corrente contesto dunque, il modello delle decision rules verrà utilizzato per osservare relazioni tra mutazioni somatiche e la tipologia tumorale
Algoritmi di Interpretable Machine Learning con applicazioni alla bioinformatica
MODOLO, RICCARDO
2022/2023
Abstract
Interpretable Machine Learning models are essential for increasing the adoption of Machine Learning within society. One of these is called decision rules, and it provides a set of rules to describe the decision criterion used by the model. CORELS is concerned with providing the optimal set of rules given a model. In this paper CORELS will be used in order to provide rules that distinguish between two tumor types based on the genetic mutations of the patients. In the current context therefore, the decision rules model will be used to observe relationships between mutations somatic and tumor typeFile | Dimensione | Formato | |
---|---|---|---|
Modolo_Riccardo.pdf
accesso aperto
Dimensione
1.17 MB
Formato
Adobe PDF
|
1.17 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/52395