作者: Mohammed Ismail Smahi , Fethellah Hadjila , Chouki Tibermacine , Mohammed Merzoug , Abdelkrim Benamar
DOI: 10.1007/978-3-319-99819-0_6
关键词:
摘要: Quality of Service (QoS) prediction is an important task in Web service selection and recommendation. Existing approaches to QoS are based on either Content Filtering or Collaborative Filtering. In the two cases, these use external data past interactions between users services predict missing future scores. One most effective techniques for Matrix Factorization (MF), with Latent Factor Models. The key idea MF consists learning a compact model both services. Thereafter simply computed as dot product user’s latent service’s model. Despite successful results recommendation area, there still set problems that should be handled, like: (i) sparsity input models, (ii) factors which prone over-fitting. this paper, we propose approach solve by using simple neural network, auto-encoder, exploiting cross-validation well-known dataset, order select ideal number factors, thereby reduce over-fitting phenomenon.