1. What is Density-Based Spatial Clustering of Applications with Noise DBSCAN (Density-based spatial clustering of applications with noise) uses density-based clustering among clustering algorithms.
1. K-Means It can be used for easy, concise, and large data. If the number of features becomes too large with distance-based algorithms, the performance of clustering is degraded. Therefore, in some c
1. What is GMM It is one of several models applying the Expectation Maximum (EM) algorithm. What is EM algorithm? EM algorithm is basically an algorithm mainly used for Unsupervised learning. It is al
1. What is K-means Clustering The K-Means clustering algorithm does not automatically identify and group the number of clusters by looking at the data. The number of clusters should be specified and t
Clustering is an example of unsupervised learning. Without any label, those with close distances in the data are classified into clusters. It is different from classification, which is supervised lea
1. Reason why you need The more input data you have on machine learning, the slower the processing. Therefore, in order to speed up the processing speed of machine learning, acceleration of learning s
1. Growth Hacking 그로스해킹(Growth Hacking)은 성장(Growth)을 위한 모든 수단(Hacking)이란 뜻으로 공격 대상의 미세한 빈틈을 찾아 해킹을 하듯이 성장을 위해 고객과 유통과정 등의 공략지점을 찾아내고 이를 적극적으로 공략하는 마케팅 방법론 브랜드, 기업, 제품 매출 증가 등을 위한 가설을 수립하고 이를 빠르게 MVP 모
Journal/Conference: InterspeechYear(published year): 2020Author: Peng Zhang, Peng Hu, Xueliang ZhangSubject: Speaker Verification Deep Embedding Learning for Text-Dependent Speaker Verification S
Journal/Conference: NIPSYear(published year): 2017Author: Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, Illia PolosukhinSubject: NLP Atte
1. DefinitionEnsemble→ 여러 예측기를 수집해서 단일 예측기 보다 더 좋은 예측기를 만드는 것. 일반적으로 앙상블 기법을 사용하면 , 예측기 하나로 훈련하였을때 보다 , 편향은 비슷하지만 분산이 줄어든다고 알려져 있다. 배깅(bagging) 원데이터 집합으로부터 크기가 같은 표본을 여러 번 단순임의 복원추출하여 각 표본(붓스트랩 표본