Notizie
Avviso di seminario: Locality Sensitive Hashing (Prof. Portinale)
Giovedi' 9/6/2016, alle ore 11.00 in sala Seminari Istituto di Informatica (stanza 192), il Prof. Luigi Portinale terra'un seminario dal titolo: Locality Sensitive Hashing: an approach to Nearest Neighbour retrieval in a Big Data setting.
Abstract: Locality Sensitive Hashing (LSH) is a technique for the indexing and retrieval of similar points in a potentially high dimensional space. It is particularly interesting in the Big Data setting, where both the number of dimensions (features) as well as the number of points (cases) may be very large.
In the lecture, after reviewing the basics of NN retrieval, some results concerning randomized algorithms for the c-approximate R-NN and R-NN reporting problems are presented, in the context of Euclidean distance over numerical features.