Open yoonhoelee opened 3 years ago
키워드: LDA, Gibbs Sampling
배운 것: LDA를 진행해서 토큰화한 자연어를 토픽으로 분류를 하는 법에 대해서 학습했다
어려웠던 부분: 토픽을 2가지로 분류를 하면 하나의 토픽은 긍정적, 하나는 부정적인 단어들로 이루어진 토픽으로 나뉠줄 알았는데, 생각만큼 잘 되지는 않는 것 같다.
더 알고 싶은 부분 / 공부하고 싶은 부분: LDA의 목적과 파이썬 환경에서 구현을 하는 것에 대해서는 이해가 되나, 아직 수학적인 부분은 이해하지 못하였다. 이 부분에 대한 보다 많은 공부가 필요할 것 같다.
느낀 점: 토픽을 나눔으로서 좀 유의미한 결과를 도출하고 싶었는데, 그러지 못해서 아쉽다.
키워드: LDA, Gibbs Sampling
배운 것: LDA를 진행해서 토큰화한 자연어를 토픽으로 분류를 하는 법에 대해서 학습했다
어려웠던 부분: 토픽을 2가지로 분류를 하면 하나의 토픽은 긍정적, 하나는 부정적인 단어들로 이루어진 토픽으로 나뉠줄 알았는데, 생각만큼 잘 되지는 않는 것 같다.
더 알고 싶은 부분 / 공부하고 싶은 부분: LDA의 목적과 파이썬 환경에서 구현을 하는 것에 대해서는 이해가 되나, 아직 수학적인 부분은 이해하지 못하였다. 이 부분에 대한 보다 많은 공부가 필요할 것 같다.
느낀 점: 토픽을 나눔으로서 좀 유의미한 결과를 도출하고 싶었는데, 그러지 못해서 아쉽다.