Positive-unlabeled learning is an essential problem in many real-world applications with only labeled positive and unlabeled data, especially when the negative samples are difficult to identify. Most existing positive-unlabeled methods will inevitably overfit class some extent due existence of unidentified samples. This paper first analyzes overfitting proposes bound generalization errors via W...

in this note first we define the notions of fuzzy positive implicativehyper bck-ideals of types 1,2,3 and 4. then we prove some theorems whichcharacterize the above notions according to the level subsets. also we obtainthe relationships among these notions, fuzzy (strong, weak, reflexive) hyperbck-ideals and fuzzy positive implicative hyper bck-ideals of types 5,6,7and 8. then, we define the no...

Sharing Positive Experience adalah berbagi cerita tentang pengalaman positif individu kepada lain. Ketika menceritakan kembali positif, memori kejadian tersebut menjadi "stimuli" yang akan mengaktifkan positive affect pada individu. ini memiliki dampak baik secara langsung maupun tidak terhadap beberapa aspek di organisasi seperti kinerja karyawan, meminimalisir tingkat stres organisasi, serta ...