- Preprocessing and Exploratory Data Analysis
- What was used and how?
- R, dplyr, caret, ggplot2, base R, NBClust
- Results from Unsupervised Learning
June 5, 2018
Outline
Preprocessing
- 3601 observations with 8 Predictor Variables grouped by subject id, 103 subjects.
- No Response Variable
| Variable | Issue |
|---|---|
| X2 | Same Information for each observation |
| X3 (Dates) | X3 gives the same information as X7 |
| X4 and X6 | Highly Correlated with each other. One removed |
- X8 is mostly 0's, but is highly correlated with 01897 and a few others and so will be left in for analysis
- O1918 and O1940 have 33 obs while all others have 35
Exploratory Data Analysis
Exploratory Data Analysis
2. Unsupervised Learning Algorithms
| Clustering | How they Work |
|---|---|
| Hclust: Complete Linkage | Maximal intercluster dissimilarity. Compute all pairwise dissimilarities between the observations in cluster A and the observations in cluster B, and record the largest of these dissimilarities |
| Hclust: Average Linkage | Mean intercluster dissimilarity. Compute all pairwise dissimilarities between the observations in cluster A and the observations in cluster B, and record the average of these dissimilarities |
2. Unsupervised Learning Algorithms
2. Unsupervised Learning Algorithms
