# Search Results

Results tagged with Search options user 150611
11 results

Use this tag for any *on-topic* question that (a) involves R either as a critical part of the question or expected answer, & (b) is not *just* about how to use R.

Suppose you have a n-inputs 1-output useful predictor that predicts $P(Y=1|X)$. Logistic regression or random forest for example. One possibility to implement it for a p-output $Y=(Y_1,Y_2,...Y_p)$ i …
answered Oct 13 '17 by Benoit Sanchez
It is possible to use only the dissimilarity matrix for clustering (without the original points), in a way that close to Kmeans. It's called "kernel Kmeans" and is very similar to "spectral clustering …
answered Sep 13 '17 by Benoit Sanchez
The general idea to compute p-values on finite possibilities is to run over all possibilities and count how many time the statistics is greater than your observed statistics. Example for a binomial: …
answered Dec 21 '17 by Benoit Sanchez
variables. You may also provide your own preferred weights. Note that R function kNN() does it for you : https://www.rdocumentation.org/packages/DMwR/versions/0.4.1/topics/kNN As a first attempt, just use …
answered Mar 31 '17 by Benoit Sanchez
I'm not sure I'm going to answer your specific problem exactly, but most often the issue with normality testing is the following. The Shapiro-Wilk test does not test if your distribution is approxima …
answered Aug 16 '17 by Benoit Sanchez
Yes, you can use dummy coding. Imagine the factor (categorical) variable $X$ has $N$ possible values: 1,2,3,, N. Then transform it into a vector of $N$ variables where all of them are 0 except the c …
answered Dec 12 '17 by Benoit Sanchez
Everything is correct. Except that the phrase "the probability for a ball to be placed in the bin A is 30% higher than in the bin B" is ambiguous. It could mean either $P(A)-P(B)=0.3$ or $P(A)=1.3P( … answered Feb 11 '18 by Benoit Sanchez Causation is not in the data and cannot be. Data only contains correlation. Most simply, if a variable$Y$is correlated to$X$,$X$can be seen as a cause of$Y$if$X$is controlled freely by the e … answered Dec 20 '17 by Benoit Sanchez A) how I can see (plot) the bold-faced statement above (e.g., in R)? Draw values of$\sigma$on a logarithmic scale: ---- 0.01 ------------ 0.1 ----------- 1 ------------ 10 ---------- 100 … answered Feb 4 '18 by Benoit Sanchez If the drift changes over time, then you can use double exponential smoothing known as the Holt-Winters procedure. see: https://en.wikipedia.org/wiki/Exponential_smoothing#Double_exponential_smoothing R has an implementation. … answered Oct 9 '17 by Benoit Sanchez They are very similar. In both models, the probability that$Y=1$given$X$can be seen as the probability that a random hidden variable$S\$ (with a certain fixed distribution) is below a certain thr …
answered Jun 10 '17 by Benoit Sanchez