Skip to Main content Skip to Navigation
Theses

Multiple instance learning for sequence data : Application on bacterial ionizing radiation resistance prediction

Abstract : In Multiple Instance Learning (MIL) problem for sequence data, the instances inside the bags aresequences. In some real world applications such as bioinformatics, comparing a random couple ofsequences makes no sense. In fact, each instance may have structural and/or functional relationshipwith instances of other bags. Thus, the classification task should take into account this across bagrelationship. In this thesis, we present two novel MIL approaches for sequence data classificationnamed ABClass and ABSim. ABClass extracts motifs from related instances and use them to encodesequences. A discriminative classifier is then applied to compute a partial classification result for eachset of related sequences. ABSim uses a similarity measure to discriminate the related instances andto compute a scores matrix. For both approaches, an aggregation method is applied in order togenerate the final classification result. We applied both approaches to the problem of bacterialionizing radiation resistance prediction. The experimental results were satisfactory.
Complete list of metadatas

Cited literature [133 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02611719
Contributor : Abes Star :  Contact
Submitted on : Monday, May 18, 2020 - 4:01:09 PM
Last modification on : Tuesday, May 26, 2020 - 7:29:50 AM

File

2019CLFAC078_ZOGHLAMI.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02611719, version 1

Collections

Citation

Manel Zoghlami. Multiple instance learning for sequence data : Application on bacterial ionizing radiation resistance prediction. Machine Learning [cs.LG]. Université Clermont Auvergne; Université de Tunis El Manar, 2019. English. ⟨NNT : 2019CLFAC078⟩. ⟨tel-02611719⟩

Share

Metrics

Record views

41

Files downloads

24