<div dir="ltr">Dear Shingo,<div><br></div><div>This is not a published method, but it's such a simple idea that it does not require publication to be justified.</div><div>You are replacing one open parameter ('How many clusters do we want to generate? We determined to be 12') with another open parameter ('How many unique subjects do we want to include in a cluster? We determined 80%'). The latter has more intuitive value than the former one.</div><div><br></div><div>In my grroupSIFT() toolbox (still in alpha version), I use this 'minimum number of unique subject' criterion to threshold connectivity edges. Even if you find an interesting looking result, if it is from only < 30 % of subjects it's difficult to be justified.</div><div><br></div><div>Which reminds me that in one of Scott's previous works (Onton and Makeig, 2009), they captured only 35% of total variance--not due to cleaning, but for dimension reduction. Probably this is the record of the smallest variance analyzed in the history of EEG study.</div><div><br></div><div>Makoto</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Nov 29, 2017 at 3:14 PM, 時本真吾 <span dir="ltr"><<a href="mailto:tokimoto@mejiro.ac.jp" target="_blank">tokimoto@mejiro.ac.jp</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Makoto,<br>
<br>
Thank you for the quick and kind reply. This is really practical and easy to understand. Is this method published anywhere? I am referring to Takahashi & Kitazawa (2017, JNS) for Silhouette value as a criterion of IC clustering.<br>
<span class=""><br>
******************************<wbr>************<br>
Shingo Tokimoto, Ph.D.<br>
in Linguistics and Psychology<br>
Department of Foreign Languages<br>
Mejiro University<br>
4-31-1, Naka-Ochiai, Shinjuku, Tokyo,<br>
161-8539, Japan<br>
<a href="mailto:tokimoto@mejiro.ac.jp">tokimoto@mejiro.ac.jp</a><br>
******************************<wbr>************<br>
<br>
</span>> 2017/11/29 3:04、Makoto Miyakoshi <<a href="mailto:mmiyakoshi@ucsd.edu">mmiyakoshi@ucsd.edu</a>>のメール:<br>
<div class="HOEnZb"><div class="h5">><br>
> Dear Shingo,<br>
><br>
> Another practical criterion to determine the number of clusters is to specify the minimum number of unique subjects per cluster. For example, if you determine to include at least 70% of unique subjects, after clustering you count how many unique subjects are included, and if it did not reach 70%, you lower the number of clusters and try it again until it meets the 70%-criterion. In most of my plugins for STUDY anlaysis, it is reported how many unique subjects are included by each cluster.<br>
><br>
> Makoto<br>
><br>
> On Sat, Nov 25, 2017 at 2:12 AM, 時本真吾 <<a href="mailto:tokimoto@mejiro.ac.jp">tokimoto@mejiro.ac.jp</a>> wrote:<br>
> Dear Dr. Jens Bernhardsson,<br>
><br>
> I am deeply grateful for the excellent code. I am very sorry for the late response. This is because I took some time to understand the meaning of plotted results. I am trying to cluster 323 ICs from 31 participants by dipoles. I repeated the code 10 times, and the number of clusters for which the average Silhouette value was greatest was 15 or 16. Can I understand that the best number of clusters here is 15 or 16? Thank you very much again.<br>
><br>
> ******************************<wbr>************<br>
> Shingo Tokimoto, Ph.D.<br>
> in Linguistics and Psychology<br>
> Department of Foreign Languages<br>
> Mejiro University<br>
> 4-31-1, Naka-Ochiai, Shinjuku, Tokyo,<br>
> 161-8539, Japan<br>
> <a href="mailto:tokimoto@mejiro.ac.jp">tokimoto@mejiro.ac.jp</a><br>
> ******************************<wbr>************<br>
><br>
> > 2017/11/17 19:19、Bernhardsson Jens <<a href="mailto:Jens.Bernhardsson@miun.se">Jens.Bernhardsson@miun.se</a>><wbr>のメール:<br>
> ><br>
> > Hi,<br>
> ><br>
> > I believe that the Silhouette value found in A10: MI-clust is based on kmeans cluster of multidimensional scaled data performed on the mutual information matrix and the kmeans found in STUDY is based on a pca derived (pre clustering) array. You can run the code below after you pre cluster in STUDY to get the mean and individual Silhouette coefficients for different cluster solutions.<br>
> ><br>
> > avgSil = [];<br>
> > kmod = [];<br>
> > siSil = [];<br>
> ><br>
> > for num_of_cluster = 2:25<br>
> > kmod = kmeans(STUDY.etc.preclust.<wbr>preclustdata,num_of_cluster, 'replicates',10,'emptyaction',<wbr>'drop','distance','<wbr>sqEuclidean');<br>
> > siSil(:,num_of_cluster-1) = silhouette(STUDY.etc.preclust.<wbr>preclustdata,kmod,'<wbr>sqEuclidean');<br>
> > avgSil = [avgSil; num_of_cluster mean(siSil(:,num_of_cluster-1)<wbr>)];<br>
> > end<br>
> ><br>
> > figure;<br>
> > plot( avgSil(:,1),avgSil(:,2),'r*-.'<wbr>);<br>
> > % ylim([0 1])<br>
> > set(gca,'XTick',1:num_of_<wbr>cluster, 'XGrid', 'on');<br>
> ><br>
> ><br>
> > Regards,<br>
> > Jens<br>
> ><br>
> > -----Ursprungligt meddelande-----<br>
> > Från: eeglablist [mailto:<a href="mailto:eeglablist-bounces@sccn.ucsd.edu">eeglablist-bounces@<wbr>sccn.ucsd.edu</a>] För ????<br>
> > Skickat: den 17 september 2017 15:43<br>
> > Till: <a href="mailto:eeglablist@sccn.ucsd.edu">eeglablist@sccn.ucsd.edu</a><br>
> > Ämne: [Eeglablist] Silhouette value for IC clustering<br>
> ><br>
> > Dear EEGLAB users,<br>
> ><br>
> > I understand that the determination of the number of clusters in IC clustering is one of the difficult problems in source localizations by ICs. Silhouette value is one of the criteria for the best number of clusters in IC clustering, and I have found the description of silhouette value in EEGLAB Wiki (A10: MI-clust) for a single participant. Can we get silhouette values for group data in STUDY formats? Thank you in advance.<br>
> ><br>
> > ******************************<wbr>************<br>
> > Shingo Tokimoto, Ph.D.<br>
> > in Linguistics and Psychology<br>
> > Department of Foreign Languages<br>
> > Mejiro University<br>
> > 4-31-1, Naka-Ochiai, Shinjuku, Tokyo,<br>
> > 161-8539, Japan<br>
> > <a href="mailto:tokimoto@mejiro.ac.jp">tokimoto@mejiro.ac.jp</a><br>
> > ******************************<wbr>************<br>
> ><br>
> > ______________________________<wbr>_________________<br>
> > Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" rel="noreferrer" target="_blank">http://sccn.ucsd.edu/eeglab/<wbr>eeglabmail.html</a><br>
> > To unsubscribe, send an empty email to <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu">eeglablist-unsubscribe@sccn.<wbr>ucsd.edu</a><br>
> > For digest mode, send an email with the subject "set digest mime" to <a href="mailto:eeglablist-request@sccn.ucsd.edu">eeglablist-request@sccn.ucsd.<wbr>edu</a><br>
><br>
> ______________________________<wbr>_________________<br>
> Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" rel="noreferrer" target="_blank">http://sccn.ucsd.edu/eeglab/<wbr>eeglabmail.html</a><br>
> To unsubscribe, send an empty email to <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu">eeglablist-unsubscribe@sccn.<wbr>ucsd.edu</a><br>
> For digest mode, send an email with the subject "set digest mime" to <a href="mailto:eeglablist-request@sccn.ucsd.edu">eeglablist-request@sccn.ucsd.<wbr>edu</a><br>
><br>
><br>
><br>
> --<br>
> Makoto Miyakoshi<br>
> Swartz Center for Computational Neuroscience<br>
> Institute for Neural Computation, University of California San Diego<br>
<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr">Makoto Miyakoshi<br>Swartz Center for Computational Neuroscience<br>Institute for Neural Computation, University of California San Diego<br></div></div>
</div>