데이터 사이언스

Netflix PrizeRecommendation and Collaborative FilteringKNN-based MethodsMartrix FactorizationRecent RecommendersCase StudyNetflix PrizeRMSE(Root Mean Square Error)$\frac{1}{\left|R \right|}\sqrt{\sum_{(i,x)}^{}(\hat{r}_{xi}-r_{xi})^2}$넷플릭스 원래 RMSE 인 0.9514에서 10% 더 줄여라!Recommendation and Collaborative Filtering추천 알고리즘의 종류User Profile matchingContent based recommendationCollaborative filtering(usi..

What is cluster analysisCategories & Basic Concepts of ClusteringPartitioning MethodsHierarchical Methods⬇️ 여기부터Integration of Hierarchical & Distance-based ClusteringDensity Based MethodsSummaryIntegration of Hierarchical & Distance-based Clusteringhierarchical clustering 은 데이터가 커짐에 따라 너무 시간 복잡도가 커져버림, 차라리 K-means 쓰는 것이 나을 정도BIRCH(Balanced Iterative Reducing and Clustering using Hierarchies)Pha..

What is cluster analysisCategories & Basic Concepts of ClusteringPartitioning MethodsHierarchical Methods⬇️ 여기부터Integration of Hierarchical & Distance-based ClusteringDensity Based MethodsSummaryIntegration of Hierarchical & Distance-based Clusteringhierarchical clustering 은 데이터가 커짐에 따라 너무 시간 복잡도가 커져버림, 차라리 K-means 쓰는 것이 나을 정도BIRCH(Balanced Iterative Reducing and Clustering using Hierarchies)Pha..

Getting to know Your DataData objects and Feature TypesNominal - {red, blue, white,... } Binary - 0, 1Ordinal - {small, medium, large}Numeric Ratio-scaledInterval-scaledBasic Statistical Description of DataMeanMedianMean과 달리 데이터가 추가되면 다시 계산해야 함 ➡️ interpolation으로 해결Mode - 가장 자주 나타나는 valueSymmetric vs. Skewed Data - 대칭 또는 비대칭(치우쳐진) 데이터Quartiles, outliers and boxplotsIQR(Inter-quartile range)min, ..

What is cluster analysisCategories & Basic Concepts of ClusteringPartitioning MethodsHierarchical Methods⬆️ 여기까지Integration of Hierarchical & Distance-based ClusteringDensity Based MethodsSummaryWhat is cluster analysisCluster: 같은 cluster 안의 데이터는 유사하다Cluster Analysis: data 사이의 유사함을 찾아내는 것Unsupervised learning유사도는 distance function으로 알아냄Good clustering ➡️ cluster 안의 유사도가 높음Categories & Basic Conc..

What is cluster analysisCategories & Basic Concepts of ClusteringPartitioning MethodsHierarchical Methods⬆️ 여기까지Integration of Hierarchical & Distance-based ClusteringDensity Based MethodsSummaryWhat is cluster analysisCluster: 같은 cluster 안의 데이터는 유사하다Unsupervised learningGood clustering ➡️ cluster 안의 유사도가 높음Categories & Basic Concepts of ClusteringMajor clustering ApproachesPartitioning approach..

시험 준비용 자료이다 어떻게 공부할까 하다가제일 간단한 키워드 부터 시작해서 3단계로 점점 살을 붙여나갈거다!!계속 반복해서 보면서 학습하면 좋을 듯 하다. 가보자고What is cluster analysisCategories & Basic Concepts of ClusteringPartitioning MethodsHierarchical MethodsIntegration of Hierarchical & Distance-based ClusteringDensity Based MethodsSummaryWhat is cluster analysisClusterPartitioning MethodsK-Means ClusteringK-modesK-MedoidsPAMCLARA(Clustering Large Applica..