How to name clusters in cluster analysis Videos

Hi TayeenulIn case of several variables, PCA is somehow essential to isolate the variables having high association among them and so it is manageable to work with few attributes with out any loss of information.Based on your example of 4 (exogenous / endogenous) categorical variable with 20 items each (record), even if it is highly correlated you could do clustering without the application of PCA.

Is it necessary to conduct CATPCA (categorical principal component analysis) prior to starting the cluster analysis, and can you please tell me after CATPCA how can I proceed for cluster analysis apparently the method. As I have four exogenous variable which contains 20 items.

One option is to treat Likert data as equal interval data, in which case you can use K-means using any distance measure (Euclidean, for example).CheersJigsaw Faculty Team

...for the cases used to derive the classification rule. The goal of cluster analysis is to identify the actual groups. For example, if you are interested in distinguishing between several disease groups using discriminant analysis, cases with known diagnoses must be available. Based on these cases, you derive a rule for classifying undiagnosed patients. In cluster analysis, you don't know who or what belongs in which group. You often don't even know the number of groups.". Hope it helps.

Nice vid StatArena! Added to "Statistics Videos" & Featured CHs: youtube.com/channel/UCJcBCQAwoNOWueJYpHc8RWA .Im Psychologist/Statistician (Msc).I do Statistical Analysis using PSPP (sim. to SPSS), R (lavaan,semtools,semplots, sim. to AMOS).Im Writer of statistical articles: "estatistics.eu/articles". You may check my stuff too and LiCS. U can reach me By email: [email protected] or by Face: EstatisticsEU LiCS: Like, Comment, Share/Subscribe

nice,,,, just this query,,,, can respondents in a cluster be electronically separated and analysed further as separate group ? I come form health industry,,,, here we have flow cytometry data GATED on scatterplots in specialized softwares,,,, can SPSS do the same to select a cluster for subsequent analysis.?

You cluster, say, respondents on the basis of something (e.g. height, income, gender etc). These become the variables and bases for clustering. For answer to second part of your question, please watch at 2.19. Thanks for watching and commenting.

Cluster analysis can create a new nominal variable for the clusters (as shown in the video you watched). To analyze each group further, you could either use the "select cases" or the split file option in SPSS under "Data" menu.

Unless "Classify" is known by some other label in version 21, your SPSS program/software seems to be incomplete (i.e. it does not include all the modules). Can not comment for sure as I'm yet to use v.21. Thank you.

In SPSS, how do you generate the BIC and log likelihood values for two-step cluster analysis? I see these reported in papers as indices of fit

great video and you have my like but I was wondering how do you choose the variables for clustering? Do they always have to be scalar?

Wa'alaikum asalaam akhi. I'm doing an SPSS course on Market segmentation. Very helpful video indeed.

Very well done sir! I was looking for a good video on this topic for some time now. Thanks !

When I click on Analyze, I do not get all those options such as Classify. I have SPSS 21.

+sentdex actually not getting any perfect tutorial like u. That's y i asked can u give me some links plz if possible

+Sagnik Saha Probably still a few months away from revisiting ML. There are tons of resources for learning ML though, doesn't have to be me. Hit up Google search.

+Fikre Beyene google "choosing the best estimator" and find the image from scikit learn. It's like a flowchart of what to choose under various circumstances.

+Sheli Ostrow Definitely, PCA is a must have topic. Not sure when, but it's on the list.

plzzz reply, I have a module depending on this

+sentdex would you kindly send me the dataset at [email protected], if you can find it that is. The link for the dataset is broken and after hours of searching on the internet for a something similar, I couldn't find anything.

I have it somewhere, definitely backed up in a few places. I wouldn't get rid of a dataset like that one. It was really good, lots of tricky lines. Was actually planning to find it again to use in some of these videos with NLTK. I do not know where you can find another labeled dataset like it, but schools tend to host all kinds of epic data sets. You can email me if you want, I'll dig it up possibly. It's definitely buried deep, but I am sure I still have it.

+sentdex Tks mate, have done already, the Udacity Nanodegree(200/month) and some others courses at Edx and Coursera too, none have the same approach as you, very practical and simple. well hope you keep doing, and I will donate for sure!

I don't want to have a paywall for content. Thought about indigogo/kickstarters for specific niche series in the past, but meh, they take too much time to set up and promote, rather just spend that time teaching the content. I may have something more specific for putting money towards specific development series via pythonprogramming.net, currently re writing that, almost done with the porting.You're free to show your support for specific content via donations at the moment, that's about it. I have all the stats for this series, and I can compare interest at release and through time to any of the others.Next topic will be Kivy. I wont be leaving scikit-learn behind for good. I am not too worried about someone else doing a series on the same topic, I'd welcome it. It'll probably take anyone who hasn't already started releasing content longer to get to this point than it will for me to do kivy and come back.We need more people covering these topics anyways. I could easily have an entire year dedicated to scikit-learn and machine learning.Stay tuned.Coursera has a nice free machine learning course. Covers a lot of really great stuff. If you need your ML fix, head there :P

+sentdex By "open source" I meant software like ruby, python, and rails ... they have attorneys and avoid any software that involves payment.  Although they do pay for support on some software, like elasticsearch ... a common approach.  I'm not sure I share your confidence about getting this to work in python/scikit, but I'm going to give a try.

many companies require open source software, so they can run code in-house. I wonder what you mean by "open source," and if they know that the traditional "open source" license actually may still be a paid license, especially when used commercially. What you want to do is certainly do-able, and within the scope of something easily done with very little customization from scikit-learn

+Chris Smith You should go at it yourself! Just take it one step at a time. Size of data shouldn't change things too much, besides possibly locking out some algos, or processing time. How big are we talking.

+sentdex Thanks, that looks on target.  I know how to gather network data, but still need to figure out how to train or how much to train, and how/when to spot anomalies ... the scale of the data is just a bit intimidating.  Hopefully I can find some "beginner" code, or just start with what I've learned from these videos.

//scikit-learn.org/stable/modules/outlier_detection.html

hi, could you tell me which tool/software you use to make the video?

PASW earlier called SPSS made by IBM corporation

Sorry for the slow response. I used Camtasia.

00:45 is that really the right name?

Do you mean you specify the number of elements of the cluster or just that they are (each) constant?

Thank you for all your tutorials so far. They are great basics for LabVIEW learners.

The constant tells the cluster building icon what the cluster will be composed of.

I will have to look into this, but I'm sure I can get something together.

Hello What's that constant (at 3:20) for? regards

I mis-numbered them. This should be #7...

@shenaruto Thanks for watching

where is tutorial 7?

OK, thanks

Thanks!

+sk "For performance scalability, each Borgmaster replica runs a stateless link shard to handle the communication with some of the Borglets; the partitioning is recalculated whenever a Borgmaster election occurs. For resiliency, the Borglet always reports its full state, but the link shards aggregate and compress this information by reporting only differences to the state machines, to reduce the update load at the elected master. If a Borglet does not respond to several poll messages its machine is marked as down and any tasks it was running are rescheduled on other machines. If communication is restored the Borgmaster tells the Borglet to kill those tasks that have been rescheduled, to avoid duplicates. A Borglet continues normal operation even if it loses contact with the Borgmaster, so currently-running tasks and services stay up even if all Borgmaster replicas fail."