Abstract | ||
---|---|---|
Classification is a common task in Machine Learning and Data Mining. Some classification problems need to take into account a hierarchical taxonomy establishing an order between involved classes and are called hierarchical classification problems. The protein function prediction can be considered a hierarchical classification problem because their functions may be arranged in a hierarchical taxonomy of classes. This paper presents an algorithm for hierarchical classification using a centroid-based approach with two versions named HCCS and HCCSic respectively. Centroid-based techniques have been widely used to text classification and in this work we explore itâs adoption to a hierarchical classification scenario. The proposed algorithm was evaluated in eight real datasets and compared against two other recent algorithms from the literature. Preliminary results showed that the proposed approach is an alternative for hierarchical classification, having as main advantage the simplicity and low computational complexity with good accuracy. |
Year | DOI | Venue |
---|---|---|
2015 | 10.5220/0005339000250033 | ICEIS (3-1) |
Field | DocType | Citations |
Data mining,One-class classification,Computer science,Artificial intelligence,Protein function prediction,Machine learning,Centroid,Computational complexity theory,Multiclass classification | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mauri Ferrandin | 1 | 0 | 0.68 |
Fabrício Enembreck | 2 | 274 | 38.42 |
Júlio César Nievola | 3 | 0 | 1.69 |
Edson Emílio Scalabrin | 4 | 36 | 14.52 |
Bráulio Coelho Ávila | 5 | 22 | 10.63 |