Search Results

Technically, for cluster analysis you use unsupervised learning - no dependencies of $X$s, just distancies are taken into account. … If, on the other hand, you just need to cluster your data - you would prefer 2nd approach - you will just get (if visualized in plot) clusters of your data, while in 1st case you will get clusters of dependencies …

Among the methods mentioned in the DeMatteis et al. report (pp.31-48), here are some that might be of interest to you: Focused reinterviews of a sample of respondents. … Cluster analysis, and data reduction methods (principal component analysis, multiple correspondence analysis, etc.), to identify similarities between records or between some interviewers, which depending …

I am interested in understanding how a suite of response variables (soil metrics such as pH, element concentrations, rates of decomposition/minderalization, etc.) vary with 1) soil depth, and 2) time since … Is some form of penalized regression (lasso, ridge, elastic net, etc.) the best way to analyze this kind of dataset? If not, how should I proceed? …

I recommend just doing variable clustering of the set of variables. … You can request a special binary variable similarity measure be used instead related to Pr(A and B) - Pr(A)Pr(B). See this and this. …

I'm looking at quantifying similarity between sites by abundance in a heatmap (and then clustering the variable site), but only for a single species (over time). … My question is, could I still use this if I'm summing the rows of abundance over time and not multiple species (the intended use of the equation)? …

I think the other user provided a great response since you were interested in specific features of those in 1 class, etc. But since you asked about "similarity", maybe this is of interest too. … You can calculate gower distance using the cluster package. …

A bit late, but Dynamic Time Warping may be a good fit here for identifying similarity of these patterns. Here's a paper where they use it for water usage recorded by a smart meter: Steffelbauer, D. … Journal of Water Resources Planning and Management, 147(6), 04021026. https://arxiv.org/pdf/2112.13778 …

I seek a single-number statistic that resumes spatial similarity, regardless of the actual group labels which often do not correspond between the different clustering (i.e., cluster A of clustering 1 does … not necessarily correspond to cluster A of clustering 2). …

Following that, we get the PC score and then cluster the scores. Logistic PCA is implemented in the R package logisticPCA if you want to try it yourself. … Some of the obvious one are the Cosine similarity and Jaccard similarity (or Gower's distance for mixed type data). …

you that your result was actually computed from random numbers and is totally useless as a measure of similarity. … If you have some kind of sample labeling, you could compute some clustering metrics like the silhouette, leveraging the cluster assumption that samples of the same class should be "more similar" than samples …

Do I have any options for validating that my similarity scores are accurate? … I've considered using clustering algorithms to get an alternative view of which vectors are similar/dissimilar, but because something like KNN also uses distance metrics I don't feel like I'd actually …

While there are no perfect match among the randomly sampled images, at first glance it looks like there are perhaps 4 main clusters of quite similar images, with some variations inside each cluster. … Depending on your use case, this is something you may want to investigate further perhaps by testing other letters, using other metrics for similarity, or using other methods for detecting clusters. …

If a better ordering doesn't exist, as may be the case for names of salesmen, you can hierarchically cluster the salesmen and use the ladderized dendrogram to bring similar salesmen close to each other … It works nicely when you have more than a few groups so for the sake of example I'm going to duplicate the OP's dataset to have 12 salesmen: library(data.table) library(ggplot2) library(dendextend) set.seed …

Hence, we are working in a space where the vast majority of it (see below) does not actually have any meaning for our universe of objects. … There is a link with the curse of dimensionality and distances in high-dimensional space, especially relating to the vastness of such spaces. …

To evaluate the quality of the clusterings and select the optimal number of clusters, I am considering using the silhouette score. … When comparing the clusterings produced by different algorithms that use different similarity measures (e.g. …