Мини-курс МЛ ТИ: Григорий Кучеров "Hash-based data structures"
Видеозапись лекций мини-курса.
День 1.
День 2.
Григорий Кучеров (CNRS/Université Gustave Eiffel, France) прочитает мини-курс "Hash-based data structures".
Hash tables are a fundamental technique in software engineering. Beyond this primary application, hash functions have numerous other applications. In these lectures, I will overview several probabilistic data structures relying on hash functions. After summarizing common facts about hashing, we will focus on approximate membership data structures: Bloom filter and Cuckoo filter. Then we will talk about extensions of Bloom filters applied to approximate counting (Count-Min sketch) and computing frequent elements in a stream (heavy hitters). Still in the streaming setting, we will talk about the cardinality estimation problem (a.k.a. count-distinct problem). We will end by a lecture on a special type of hashing increasingly applied in practice: locality-sensitive hashing.
The lectures are oriented to bachelor/master/PhD students with initial background in algorithms and data structures and discrete probability. The lectures can be in english or russian, to be defined with the audience.
Место проведения: Покровский бульвар, 11, аудитория R405.
Язык: английский/русский
Расписание:
среда 24.11.2021 13:00-14:20
среда 24.11.2021 14:40-16:00
четверг 25.11.2021 13:00-14:20
четверг 25.11.2021 14:40-16:00 дополнительная лекция
Для прохода на лекцию необходимо иметь при себе QR-код. Онлайн-трансляция пройдет в Zoom.
Если Вы планируете посетить лекции в аудитории или присоединится к нам онлайн, пожалуйста, пройдите регистрацию.