A Practical Approach to Novel Class Discovery in Tabular Data: Full training procedure
by Dataology: Study of Data in Computer ScienceMay 26th, 2024
Too Long; Didn't Read
The complete training procedure for Novel Class Discovery (NCD) integrates models, hyperparameter optimization, and novel class estimation through k-fold Cross-Validation. Algorithm 1 provides a simplified overview of this complex process for better understanding.
Authors:
(1) Troisemaine Colin, Department of Computer Science, IMT Atlantique, Brest, France., and Orange Labs, Lannion, France;
(2) Reiffers-Masson Alexandre, Department of Computer Science, IMT Atlantique, Brest, France.;
(3) Gosselin Stephane, Orange Labs, Lannion, France;
(4) Lemaire Vincent, Orange Labs, Lannion, France;
(5) Vaton Sandrine, Department of Computer Science, IMT Atlantique, Brest, France.
Table of Links
Estimating the number of novel classes
Appendix A: Additional result metrics
Appendix C: Cluster Validity Indices numerical results
Appendix D: NCD k-means centroids convergence study
6 Full training procedure
In the previous sections, we presented the models, the hyperparameter optimization and the estimation procedure of the number of novel classes independently. In this section, these components are brought together to form a complete training procedure. To ensure that no prior knowledge about the novel classes is ever used in this process, the number of novel classes is naturally estimated during the k-fold CV introduced in Section 4. As the whole process is quite complex, we try to summarize it in clear terms in this section and in Algorithm 1.
This paper is available on arxiv under CC 4.0 license.
L O A D I N G
. . . comments & more!
. . . comments & more!
