Re: [math] Discuss: New feature MiniBatchKMeansClusterer

2020-02-26 Thread CT
Hi Gilles, -- Original -- From: "GillesSadowski"

Re: [math] Discuss: New feature MiniBatchKMeansClusterer

2020-02-25 Thread CT
Hi Gilles, -- Original -- From: "GillesSadowski"https://github.com/apache/commons-math/pull/120 > >Sorry for seemingly nit-picking but the global issue is the same >as with PR #118: It contains too many unrelated changes. >There should be *one* PR for each batch of

Re: [math] Discuss: New feature MiniBatchKMeansClusterer

2020-02-25 Thread CT
Hi, -- Original -- From: "GillesSadowski"https://github.com/apache/commons-math/pull/120 > >Sorry for seemingly nit-picking but the global issue is the same >as with PR #118: It contains too many unrelated changes. >There should be *one* PR for each batch of signif

Re: [math] Discuss: New feature MiniBatchKMeansClusterer

2020-02-25 Thread CT
Hi Gilles:   Sorry for my unfamiliar in contribution. I started a new PR for most of your suggestion: https://github.com/apache/commons-math/pull/120 I remain have one question below: -- Original -- From: "Gilles Sadowski"https://github.com/apache/commons-math/pul

Re: [math] Discuss: New feature MiniBatchKMeansClusterer

2020-02-22 Thread CT
Hi Gilles: I really appricate for your patient to help me to solve the mail sending problem, I try to set the only setting about charset "Use Unicode" for this mail. I have created a pull request: https://github.com/apache/commons-math/pull/118 And a comparsion between KMeans and MiniBatchK

Re: [math]New feature MiniBatchKMeansClusterer

2020-02-04 Thread CT
told me which mail editor is work well with the ML. > Best regards, > Gilles -- Original -- From: "Gilles Sadowski";; Date: Jan 20, 2020 To: "Commons Developers List"; Subject: Re: [math]New feature MiniBatchKMeansClusterer Hi.

[math]New feature MiniBatchKMeansClusterer

2020-01-19 Thread CT
Hi,  In my picture search project, I need a cluster algorithm to narrow the dataset, for accelerate the search on millions of pictures.   First we use python+pytorch+kmean, with the growing data from thousands to millions, the KMeans clustering became slower and slower(seconds to minutes), then