Тёмный

Topic Models 

Jordan Boyd-Graber
Подписаться 13 тыс.
Просмотров 40 тыс.
50% 1

Clustering text into topics using latent Dirichlet allocation (version recorded for machine learning course)

Опубликовано:

 

12 сен 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 17   
@dhawalruud1
@dhawalruud1 9 лет назад
Hi JBG, Loving these videos as a part of your flipped classrooms! One request: Could you follow some naming convention for these videos? I know they're a part of 5622 but having them named linearly and bundled up as a RU-vid playlist would make them easier to find and help me (and other students) keep a track of their progress as the next time I load up the playlist, it automatically plays the next video. Thanks!
@JordanBoydGraber
@JordanBoydGraber 7 лет назад
Yeah, I'll have a better model next time I do it.
2 года назад
probably the best LDA video on youtube
@JordanBoydGraber
@JordanBoydGraber 2 года назад
I might disagree actually. I update my videos when I teach the class each year, so a later video might actually be better: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-fCmIceNqVog.html (If nothing else in terms of production value)
@ryanwesslen
@ryanwesslen 8 лет назад
Fantastic lecture on topic modeling! Thank you for posting.
@robertholder
@robertholder 5 лет назад
Very nice use of both high level intuition and concrete maths. Q - what do you dislike about gensim's implementation of LDA? How do you feel about the lda python module?
@kevind.shabahang
@kevind.shabahang 3 года назад
Very clear. Thank you!
@kevind.shabahang
@kevind.shabahang 3 года назад
lol'ed at 31:29: "We decided the system fully converged once it was time for bed"
@SuperSpiderme
@SuperSpiderme 8 лет назад
Great video! Very helpful.
@NotesFromPittsburgh
@NotesFromPittsburgh 7 лет назад
Thanks, well worth studying your explanations.
@aazqueta
@aazqueta 8 лет назад
Great explanation, congratulations! One question, why is Gensim not the best implementation?
@JordanBoydGraber
@JordanBoydGraber 8 лет назад
Well, perhaps it has improved recently, but when I recorded the lecture, it did not do hyperparameter optimization
@sounakmojumder808
@sounakmojumder808 5 лет назад
hi. thank you very much did you write i in place of w in equations?
@zhengnan5373
@zhengnan5373 7 лет назад
Thanks for the video. Once we have all the parameters, how do we make predictions about topic proportions on a new, unseen document?
@JordanBoydGraber
@JordanBoydGraber 7 лет назад
people.cs.umass.edu/~wallach/publications/wallach09evaluation.pdf
@Freshchris001
@Freshchris001 5 лет назад
Thanks for the video, it helped a lot! One thing I don't get is how you define which words belong to each topic.
@JordanBoydGraber
@JordanBoydGraber 5 лет назад
1) Each token gets assigned to a topic 2) Afterward, you can count up all the words that have been assigned to a topic. The more words, the higher it is in the ranking.
Далее
Latent Dirichlet Allocation (Part 1 of 2)
26:57
Просмотров 132 тыс.
НЕВОЗМОЖНЫЙ ЭКСПЕРИМЕНТ
00:39
Просмотров 74 тыс.
А ВЫ ЛЮБИТЕ ШКОЛУ?? #shorts
00:20
Просмотров 2,8 млн
An Introduction to Topic Modeling
26:39
Просмотров 65 тыс.
Matti Lyra - Evaluating Topic Models
45:05
Просмотров 24 тыс.
Topic Models: Gibbs Sampling (13c)
12:55
Просмотров 27 тыс.
Topic Models: Introduction
14:21
Просмотров 33 тыс.
Dirichlet Process Mixture Models and Gibbs Sampling
26:33
Is ChatGPT AI? Is it NLP? [Lecture]
4:04
Просмотров 670
Topic Modeling with Python
50:14
Просмотров 67 тыс.
LDA Topic Models
20:37
Просмотров 181 тыс.