First two chapters of Ignite ML book

[Alexey Zinoviev]

Hi, dear community, today I published two first chapters of Ignite ML books

   1. Apache Ignite ML: origins and development
   

   2. Apache Ignite ML: possible use cases, racing with Spark ML, plans for
   the future
   


If you have any questions or feedback please let me know.

P.S. I ask you to do +50 claps for my new article, if you have a medium
account, this will increase it in the search results and recommendations.

Re: Video overview of eight Ignite-specific talks scheduled for the IMCS Summit 2020

[Alexey Zinoviev]

Thanks for the video support, Denis, shared!

пт, 16 окт. 2020 г., 21:45 Denis Magda :

> Igniters,
>
> We've got eight (!!!) Ignite-specific sessions scheduled for the In-Memory
> Computing Summit 2020. I've gone ahead and produced a quick overview of
> those sessions. Check and weigh up which one to join:
> https://youtu.be/SCqd4qfBY6Q
>
> The sessions are delivered by our community folks: @Saikat Maitra
> , @Alexey Zinoviev , 
> @Stanislav
> Lukyanov , @Andrey Alexandrov
> , @Glenn Wiebe , @Greg
> Stachnick , me
>
> Support the speakers and share the video within your network.
>
> -
> Denis
>

Re: Preprocessing of data to use in Naive-Bayes

[Alexey Zinoviev]

Sorry, no update, I'm on vacation till October. But later I return to your
question, it's not trivial on preprocessing stage.



вт, 15 сент. 2020 г., 9:21 Priya Yadav :

> Hi,
>
>
>
> If there in any update, Please let me know.
>
>
>
> Thanks
>
>
>
> *From:* Priya Yadav
> *Sent:* Sunday, September 6, 2020 8:14 PM
> *To:* Alexey Zinoviev 
> *Cc:* user@ignite.apache.org
> *Subject:* RE: Preprocessing of data to use in Naive-Bayes
>
>
>
> Hi Alexey,
>
>
>
> I am stuck on the preprocessing step itself as I am not able to find any
> api which takes the sentence , reads the tokens and calculate their count
> whereas scikit-learn provides the apis out of the box.
>
>
>
> I am attaching the sample data that I need to categorize on the basis of
> user experience. Please find the python code snippet below:
>
>
>
> from sklearn.naive_bayes import MultinomialNB
>
> from sklearn.feature_extraction.text import CountVectorizer
>
> classifier = MultinomialNB();
>
> vect=CountVectorizer();
>
> counts=vect.fit_transform(["pizza was soft, very nice"," good ambience and
> excellent service","tool a long time, service needs improvement","toppings
> were very less, but bread was excellent"]) ;
>
> counts=vect.fit_transform(comment);
>
> targets = ['Good Experience','Good Experience','Bad Experience','Good
> Experience'];
>
> classifier.fit(counts,targets);
>
> predictComments = [“soft bread, nice toppings”]
>
> predictData=vect.transform(predictComments);
>
> predictions = classifier.predict(predictData)
>
> print(predictions);
>
>
>
>
>
> Thanks,
>
> Priya
>
>
>
>
>
> *From:* Alexey Zinoviev 
> *Sent:* Sunday, September 6, 2020 6:41 PM
> *To:* Igor Belyakov 
> *Cc:* user 
> *Subject:* Re: Preprocessing of data to use in Naive-Bayes
>
>
>
> Very interesting case!
>
>
>
> We have 3 different implementations for NaiveBayes algorithm
>
> https://apacheignite.readme.io/docs/naive-bayes
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__apacheignite.readme.io_docs_naive-2Dbayes=DwMFaQ=ObqWq9831a7badpzAhIKIA=qixDeHnSzhtciDY_pRHc4x12Ip0suDtJCZ5Ce1zlWfQ=s_IECR0VZUJ9ds7ehfpq8i3L0GTFiHRJ3ghViHS6dE8=oCy265A-SLfh0-HlWoiLAaoxQoXI4w6qOJ_BgZh66Dg=>
>
>
>
> I suppose that this is the best for this task
> https://apacheignite.readme.io/docs/naive-bayes#discrete-bernoulli-naive-bayes
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__apacheignite.readme.io_docs_naive-2Dbayes-23discrete-2Dbernoulli-2Dnaive-2Dbayes=DwMFaQ=ObqWq9831a7badpzAhIKIA=qixDeHnSzhtciDY_pRHc4x12Ip0suDtJCZ5Ce1zlWfQ=s_IECR0VZUJ9ds7ehfpq8i3L0GTFiHRJ3ghViHS6dE8=S0CrU7joi3OwZA5W7BunClUM8cv-m2HtQziDPhuDtlg=>
>
> Data should be prepared as Vectors in Ignite Cache to start training.
>
>
>
> Dear Priya Yadav, could you please provide code or pseudocode with how you
> populate your Ignite cache with sentences data, a few sentences will be
> useful too.
>
> Also will be useful, how could you solve this task in scikit-learn, I'll
> try to help with the preprocessing code for this case.
>
>
>
> Sincerely yours,
>
>Alexey
>
>
>
> пт, 4 сент. 2020 г. в 19:40, Igor Belyakov :
>
> Alexey,
>
>
>
> Do you have any thoughts regarding that?
>
>
>
> Igor
>
>
>
> On Fri, Sep 4, 2020 at 10:03 AM Priya Yadav  wrote:
>
> Hi,
>
>
>
> Problem Statement: I have a feedback sentences having words separated by
> spaces like normal English sentences. Using these sentences I need to
> classify into categories based on some keywords. How should I preprocess my
> data in order to use it in Naive-Bayes.
>
> Any leads would be helpful.
>
> Thanks in advance.
>
>
>
> This email and any files transmitted with it are confidential, proprietary
> and intended solely for the individual or entity to whom they are
> addressed. If you have received this email in error please delete it
> immediately.
>
> This email and any files transmitted with it are confidential, proprietary
> and intended solely for the individual or entity to whom they are
> addressed. If you have received this email in error please delete it
> immediately.
>

Re: Preprocessing of data to use in Naive-Bayes

[Alexey Zinoviev]

Very interesting case!

We have 3 different implementations for NaiveBayes algorithm
https://apacheignite.readme.io/docs/naive-bayes

I suppose that this is the best for this task
https://apacheignite.readme.io/docs/naive-bayes#discrete-bernoulli-naive-bayes
Data should be prepared as Vectors in Ignite Cache to start training.

Dear Priya Yadav, could you please provide code or pseudocode with how you
populate your Ignite cache with sentences data, a few sentences will be
useful too.
Also will be useful, how could you solve this task in scikit-learn, I'll
try to help with the preprocessing code for this case.

Sincerely yours,
   Alexey

пт, 4 сент. 2020 г. в 19:40, Igor Belyakov :

> Alexey,
>
> Do you have any thoughts regarding that?
>
> Igor
>
> On Fri, Sep 4, 2020 at 10:03 AM Priya Yadav  wrote:
>
>> Hi,
>>
>>
>>
>> Problem Statement: I have a feedback sentences having words separated by
>> spaces like normal English sentences. Using these sentences I need to
>> classify into categories based on some keywords. How should I preprocess my
>> data in order to use it in Naive-Bayes.
>>
>> Any leads would be helpful.
>>
>> Thanks in advance.
>>
>>
>> This email and any files transmitted with it are confidential,
>> proprietary and intended solely for the individual or entity to whom they
>> are addressed. If you have received this email in error please delete it
>> immediately.
>>
>

Re: Launched Ignite meetups and redesigned events pages

[Alexey Zinoviev]

Count me in!

пн, 27 янв. 2020 г. в 15:12, Dmitriy Pavlov :

> Hi Denis,
>
> yes, sorry for late reply, I just did double check that I can access
> answers. Additionally, Ksenia R. have access to proposals.
>
> Anyone from PMC who would like to volunteer and being PMC Representative
> for Apache Ignite Local meetups are always welcomed (according
> https://www.apache.org/foundation/marks/events  - The ASF events branding
> policy).
>
> Moreover, not-yet-PMC volunteers, who would help with talks preparation are
> also welcomed. I strongly believe we should not be too formal here.
>
> Sincerely,
> Dmitriy Pavlov
>
>
>
> чт, 9 янв. 2020 г. в 23:47, Denis Magda :
>
> > Dmitry,
> >
> > We've also added the reference to the form on the event's webpage. Btw,
> > could you remind me who will be receiving the proposals - you, I and some
> > other folks or all @dev list?
> >
> > -
> > Denis
> >
> >
> > On Thu, Jan 9, 2020 at 1:53 AM Dmitriy Pavlov 
> wrote:
> >
> >> Hi Igniters, Mauricio, Ignacio, and Denis,
> >>
> >> Thank you all for updating these pages.
> >>
> >> I would like to stress just one thing:
> >> should you have in mind some topic you can run talk about, please feel
> >> absolutely free to fill submission proposal.
> >>
> >> To submit proposal you need to fill in this form
> >>
> >>
> https://docs.google.com/forms/d/e/1FAIpQLSdiY7movHKvyWg3gOVedHgukJJnNiaejSO_X838vBseL9VmiQ/viewform
> >> .
> >>
> >> Sincerely,
> >> Dmitriy Pavlov
> >>
> >> вт, 7 янв. 2020 г. в 02:48, Denis Magda :
> >>
> >> > Igniters,
> >> >
> >> > I've just merged changes contributed by Mauricio and Ignacio, who
> helped
> >> > us to prepare professional webpages for Ignite meetup groups around
> the
> >> > world [1] and for Ignite events [2]. Now you can easily enroll in a
> >> meetup
> >> > group nearby or sign up for an upcoming event to be hosted one of our
> >> > experts.
> >> >
> >> > Please check them out and share your feedback. In the meantime, three
> of
> >> > us are going to carry on with optimizations and changes making the
> >> website
> >> > more useful for developers as well as more searchable/discoverable
> from
> >> > various search engines.
> >> >
> >> > [1] https://ignite.apache.org/meetup-groups.html
> >> > [2] https://ignite.apache.org/events.html
> >> >
> >> > -
> >> > Denis
> >> >
> >>
> >
>

Re: How to use spark-2.4 module in Ignite 2.8?

[Alexey Zinoviev]

Hi, I've found a bug in pom. It will be fixed on next week in master and
release branches.
Thank for the message!

ср, 18 дек. 2019 г. в 22:11, Denis Magda :

> Alexey, Nikolay,
>
> Could you please step in and shed some light on what has changed with
> Spark? Probably, that's an issue that needs to be addressed before the
> release.
>
> -
> Denis
>
>
> On Wed, Dec 18, 2019 at 9:26 AM Stefan Miklosovic <
> stefan.mikloso...@instaclustr.com> wrote:
>
>> Hi,
>>
>> I want to try spark-2.4 module committed recently but I do not see
>> spark-2.4 directory in optional dir after I build whole Ignite
>> distribution.
>>
>> I am building like this:
>>
>> mvn clean install -Pall-java,all-scala,licenses -DskipTests && mvn
>> initialize -Prelease
>>
>> Why is there just "spark" but not "spark-2.4" even "spark-2.4" is in
>> all-scala profile?
>>
>> Cheers
>>
>

Re: [Webinar] Data Streaming Using Apache Flink and Apache Ignite

[Alexey Zinoviev]

Great to get a video after to watch and share later, very ineresting case

вс, 8 дек. 2019 г., 20:39 Saikat Maitra :

> Hi,
>
> I will be presenting in a webinar about Data Streaming Using Apache Flink
> and Apache Ignite on Wednesday, Dec 18th.
>
>
> https://www.gridgain.com/resources/webinars/data-streaming-using-apache-flink-and-apache-ignite
>
> Apache Ignite is a powerful in-memory computing platform. The Apache
> IgniteSink streaming connector enables users to inject Flink data into the
> Ignite cache. Join Saikat Maitra to learn how to build a simple data
> streaming application using Apache Flink and Apache Ignite. This stream
> processing topology will allow data streaming in a distributed, scalable,
> and fault-tolerant manner, which can process data sets consisting of
> virtually unlimited streams of events.
>
> Please feel free to join the webinar.
>
> Regards,
> Saikat
>

Apache Ignite on Apache Con

[Alexey Zinoviev]

Hi, Igniters, I've presented a new topic "Ensembles of ML algorithms and
Distributed Online Machine Learning with Apache Ignite" on ApacheCon in
Berlin this year.

Video is here https://www.youtube.com/watch?v=3CmnV6IQtTw
Slides are here
https://speakerdeck.com/zaleslaw/ensembles-of-ml-algorithms-and-distributed-online-machine-learning-with-apache-ignite

You could find another talks here https://aceu19.apachecon.com/schedule

Re: [DISCUSS] Proposal for Ignite Extensions as a separate Bahir module or Incubator project

[Alexey Zinoviev]

Also, dear Saikat Maitra, could you please describe how you see the release
cycles in Bahir Ignite Extensions and how it be related to Ignite release,
2.9, 3.0 for example.

Thank you for your energy

ср, 23 окт. 2019 г., 8:10 Alexey Zinoviev :

> Please, give me permissions too, I'd glad to help with this modules
> migration and support part of them in future, but also we need not only
> contributor but a few Committer permissions to merge In repository in other
> side it could be very long proccess.
>
> Could you ask Bahir Community about that?
>
> ср, 23 окт. 2019 г., 2:31 Saikat Maitra :
>
>> Hi,
>>
>> I discussed with Apache Bahir community and they are interested to have
>> Apache Ignite extensions as part of Apache Bahir project.
>>
>> I have also requested for contributor access in Jira for Apache Bahir
>> project so that I can create issues and assign to myself. I can help with
>> code reviews as well.
>>
>> Also my thoughts on releases specific to dependencies for Apache Ignite is
>> to do a fast follow up release for modules based on latest Apache Ignite
>> stable release.
>>
>> Here is the email thread for reference
>> https://www.mail-archive.com/dev@bahir.apache.org/msg02703.html
>>
>> I wanted to connect and get feedback on the proposal and if we are ok to
>> move the following Apache Ignite Extensions
>>
>>
>> https://cwiki.apache.org/confluence/display/IGNITE/IEP-36%3A+Modularization#IEP-36:Modularization-IndependentIntegrations
>>
>> Regards,
>> Saikat
>>
>>
>> On Fri, Oct 18, 2019 at 9:44 PM Saikat Maitra 
>> wrote:
>>
>> > Hello,
>> >
>> > We wanted to discuss on a proposal to move and support the Apache Ignite
>> > integrations as separate Ignite Extensions as discussed here
>> >
>> http://apache-ignite-developers.2346864.n4.nabble.com/DISCUSS-Pub-Sub-Streamer-Implementation-td43944.html
>> > .
>> >
>> > The reason we wanted to move our Apache Ignite integration as  separate
>> > Extensions is this will help us to manage and maintain separate
>> lifecycle
>> > for Apache Ignite integrations.
>> >
>> > All the integrations will continue to be part of ASF and we will  keep
>> > supporting and developing in accordance with ASF vision and practices.
>> >
>> > We are considering following two choices for moving to Apache Ignite
>> > Extensions:
>> >
>> > 1. Reach out to Apache Bahir community and propose to make Ignite
>> > Extensions a separate module as part of Apache Bahir project.
>> >
>> > https://bahir.apache.org/
>> >
>> >
>> >
>> https://blogs.apache.org/foundation/entry/the_apache_software_foundation_announces96
>> >
>> >
>> > 2. Reach out to Apache Incubator community and request for a new project
>> > for Ignite Extensions.
>> >
>> > Please review and share feedback on our proposal.
>> >
>> > Warm Regards,
>> > Saikat
>> >
>>
>

Re: [DISCUSS] Proposal for Ignite Extensions as a separate Bahir module or Incubator project

[Alexey Zinoviev]

Please, give me permissions too, I'd glad to help with this modules
migration and support part of them in future, but also we need not only
contributor but a few Committer permissions to merge In repository in other
side it could be very long proccess.

Could you ask Bahir Community about that?

ср, 23 окт. 2019 г., 2:31 Saikat Maitra :

> Hi,
>
> I discussed with Apache Bahir community and they are interested to have
> Apache Ignite extensions as part of Apache Bahir project.
>
> I have also requested for contributor access in Jira for Apache Bahir
> project so that I can create issues and assign to myself. I can help with
> code reviews as well.
>
> Also my thoughts on releases specific to dependencies for Apache Ignite is
> to do a fast follow up release for modules based on latest Apache Ignite
> stable release.
>
> Here is the email thread for reference
> https://www.mail-archive.com/dev@bahir.apache.org/msg02703.html
>
> I wanted to connect and get feedback on the proposal and if we are ok to
> move the following Apache Ignite Extensions
>
>
> https://cwiki.apache.org/confluence/display/IGNITE/IEP-36%3A+Modularization#IEP-36:Modularization-IndependentIntegrations
>
> Regards,
> Saikat
>
>
> On Fri, Oct 18, 2019 at 9:44 PM Saikat Maitra 
> wrote:
>
> > Hello,
> >
> > We wanted to discuss on a proposal to move and support the Apache Ignite
> > integrations as separate Ignite Extensions as discussed here
> >
> http://apache-ignite-developers.2346864.n4.nabble.com/DISCUSS-Pub-Sub-Streamer-Implementation-td43944.html
> > .
> >
> > The reason we wanted to move our Apache Ignite integration as  separate
> > Extensions is this will help us to manage and maintain separate lifecycle
> > for Apache Ignite integrations.
> >
> > All the integrations will continue to be part of ASF and we will  keep
> > supporting and developing in accordance with ASF vision and practices.
> >
> > We are considering following two choices for moving to Apache Ignite
> > Extensions:
> >
> > 1. Reach out to Apache Bahir community and propose to make Ignite
> > Extensions a separate module as part of Apache Bahir project.
> >
> > https://bahir.apache.org/
> >
> >
> >
> https://blogs.apache.org/foundation/entry/the_apache_software_foundation_announces96
> >
> >
> > 2. Reach out to Apache Incubator community and request for a new project
> > for Ignite Extensions.
> >
> > Please review and share feedback on our proposal.
> >
> > Warm Regards,
> > Saikat
> >
>

Re: Application for Ignite Contributor

[Alexey Zinoviev]

Hi, Deepak, this is an Umbrella ticket for newbies in ML
https://issues.apache.org/jira/browse/IGNITE-9281

I'll freshen up it and add more tasks next week to have more option to
start with ML
Of course, you could free contact with me here, on dev-list, or on github
by mention @zaleslaw preparing PR for the Ignite

Good luck!

чт, 26 сент. 2019 г. в 12:38, Dmitriy Pavlov :

> Hi Deepak,
>
> Thank you for your interest in Apache Ignite.
>
> Have you ever considered the applicability of usage of Apache Ignite in
> Apache OFBiz. Would it be useful as scalable storage? As storage option?
>
> Maybe we can think about some integration of these two projects. WDYT?
>
> One more thing, do you know about static code inspections  Apache OFBiz was
> set up recently.  Was it successful?
>
> Sincerely,
> Dmitriy Pavlov
>
> ср, 25 сент. 2019 г. в 22:49, Denis Magda :
>
> > Hi Deepak and welcome!
> >
> > Sorry for the late response. You sent this message to the user list,
> > forwarded it to the dev list.
> >
> > I added you to the Ignite contributors list in JIRA. Feel free to take
> > over any ticket you like. Do you have any specific interest (SQL, ML,
> > caching, etc.?
> >
> > -
> > Denis
> >
> >
> > On Sat, Sep 21, 2019 at 2:35 AM Deepak Nigam  >
> > wrote:
> >
> >> Hello all,
> >>
> >> Please consider my application to become an Ignite Contributor. Here are
> >> the details:
> >>
> >> Full Name: Deepak Nigam
> >> ASF User Name: deepaknigam
> >> Email Address: deepaknigam.1...@gmail.com
> >>
> >> I am already a committer in the Apache OFBiz project, hence already
> >> signed ICLA.
> >>
> >> Regards,
> >> --
> >> Deepak Nigam
> >>
> >
>

Re: Professional Ignite Course

[Alexey Zinoviev]

Great course!

чт, 26 сент. 2019 г. в 03:56, sri hari kali charan Tummala <
kali.tumm...@gmail.com>:

> I did went trough the course and I asked pluralsight to add more, in the
> next course I would like to see ignite installation on aws and spark +
> ignite integration.
> Real time analytics with spark + ignite.
>
>
> On Wednesday, September 25, 2019, Denis Magda  wrote:
>
> > Igniters,
> >
> > Recently, I've come across a professionally crafted course about Apache
> > Ignite. It will be interesting for some of you:
> > https://www.pluralsight.com/courses/apache-ignite-getting-started
> >
> > Edward, thanks for the course! I've added it to Ignite Doc's menu right
> > below the book.
> >
> > -
> > Denis
> >
>
>
> --
> Thanks & Regards
> Sri Tummala
>

Re: [ANNOUNCE] Apache Ignite 2.7.6 Released

[Alexey Zinoviev]

Great, go 2.8!

пт, 20 сент. 2019 г. в 13:25, Alexey Goncharuk :

> The Apache Ignite Community is pleased to announce the release of Apache
> Ignite 2.7.6.
>
> Apache Ignite [1] is a memory-centric distributed database, caching, and
> processing platform for transactional, analytical, and streaming workloads
> delivering in-memory speeds at petabyte scale.
>
> This release introduces addresses frequent usability and critical stability
> issues
> https://ignite.apache.org/releases/2.7.6/release_notes.html
>
> Download the latest Ignite version from here:
> https://ignite.apache.org/download.cgi
>
> Please let us know [2] if you encounter any problems.
>
> Regards,
> Alexey Goncharuk on behalf of Apache Ignite Community
>
> [1] https://ignite.apache.org
> [2] https://ignite.apache.org/community/resources.html
>

Re: ML stable and performance

[Alexey Zinoviev]

I've started the discussion here
http://apache-ignite-developers.2346864.n4.nabble.com/Proposed-Release-2-7-7-2-7-6-ML-extra-fixed-bugs-td43537.html


пт, 13 сент. 2019 г. в 22:12, Alexey Zinoviev :

> The reason was that the last year there is no significant releases of
> Ignite between 2.7 and 2.8, only minor releases with long story of renaming.
> I am and another ML guys are ready in 1-2 months prepare ML module for 2.8
> or for the minor release 2.7.7 = 2.7.6 + updated ML + new fixed bugs
>
> Let's discuss it in separate thread next week
>
>
>
> пт, 13 сент. 2019 г. в 21:55, Denis Magda :
>
>> Alexey, I'm wondering,
>>
>> Are there any dependencies on Ignite Core that make us put off the ML
>> changes release until 2.8? I know that you do not support the idea of ML
>> as
>> a separate Ignite module but this concept would allow us to release ML as
>> frequently as we want not being blocked by Ignite core releases.
>>
>>
>> -
>> Denis
>>
>>
>> On Fri, Sep 13, 2019 at 11:45 AM Alexey Zinoviev 
>> wrote:
>>
>> > I could answer as one of developers of ML module.
>> > Currently is available the ML in version 2.7.5, it supports a lot of
>> > algorithms and could be used in production, but the API is not stable
>> and
>> > will be changed in 2.8
>> >
>> > The ML module will be stable since next release 2.8, also we have no
>> > performance report to compare for example with Spark ML
>> > Based on my exploration the performance of in terms of Big O notation is
>> > the same like in Spark ML (real numbers says that Ignite ML is more
>> faster
>> > than Spark ML due to Ignite in-memory nature and so on)
>> >
>> > Since 2.8 it will have good integration with TensorFlow, Spark ML,
>> XGBoost
>> > via model inference.
>> >
>> > You as a user have no ability to run, for-example scikit-learn or R
>> > packages in distributed mode over Ignite, but you could run the
>> TensorFlow,
>> > using Ignite as a distributed back-end instead of Horovod.
>> >
>> > If you have any questions, please let me know
>> >
>> >
>> >
>> > пт, 13 сент. 2019 г. в 21:28, Denis Magda :
>> >
>> >> David,
>> >>
>> >> Let me loop in Ignite dev list that has Ignite ML experts subscribed.
>> >> Please, could you share more details in regards to your performance
>> >> testing
>> >> and objectives for Ignite ML overall?
>> >>
>> >> The module is ready for production and we're ready to help solve any
>> >> cornerstones.
>> >>
>> >> -
>> >> Denis
>> >>
>> >>
>> >> On Fri, Sep 6, 2019 at 4:50 AM David Williams 
>> >> wrote:
>> >>
>> >> > Python is 25 times slower than Java for ML at runtimes, which I found
>> >> out
>> >> > online. But I don't know that statement is true or not. I need
>> insiders'
>> >> > opinion.  Which ml other packages are best options for Ignite?
>> >> >
>> >> > On Fri, Sep 6, 2019 at 7:28 PM Mikael 
>> >> wrote:
>> >> >
>> >> >> Hi!
>> >> >>
>> >> >> I have never used it myself but it's been there for long time and I
>> >> >> would expect it to be stable, and yes it will run distributed, I
>> can't
>> >> >> say anything about performance as I have never used it.
>> >> >>
>> >> >> You will find a lot of more information at:
>> >> >>
>> >> >> https://apacheignite.readme.io/docs/machine-learning
>> >> >>
>> >> >> Mikael
>> >> >>
>> >> >>
>> >> >> Den 2019-09-06 kl. 11:50, skrev David Williams:
>> >> >> >
>> >> >> >
>> >> >> > I am evaluating ML framework for Java platform. I knew Ignite has
>> ML
>> >> >> > package.
>> >> >> > But I like to know its stability and performance for production.
>> Can
>> >> >> > Ignite
>> >> >> > ML code run in distribute way?
>> >> >> >
>> >> >> > Except its own ML package, which ml packages are best options for
>> >> >> Ignite?
>> >> >>
>> >> >
>> >>
>> >
>>
>

Re: ML stable and performance

[Alexey Zinoviev]

The reason was that the last year there is no significant releases of
Ignite between 2.7 and 2.8, only minor releases with long story of renaming.
I am and another ML guys are ready in 1-2 months prepare ML module for 2.8
or for the minor release 2.7.7 = 2.7.6 + updated ML + new fixed bugs

Let's discuss it in separate thread next week



пт, 13 сент. 2019 г. в 21:55, Denis Magda :

> Alexey, I'm wondering,
>
> Are there any dependencies on Ignite Core that make us put off the ML
> changes release until 2.8? I know that you do not support the idea of ML as
> a separate Ignite module but this concept would allow us to release ML as
> frequently as we want not being blocked by Ignite core releases.
>
>
> -
> Denis
>
>
> On Fri, Sep 13, 2019 at 11:45 AM Alexey Zinoviev 
> wrote:
>
> > I could answer as one of developers of ML module.
> > Currently is available the ML in version 2.7.5, it supports a lot of
> > algorithms and could be used in production, but the API is not stable and
> > will be changed in 2.8
> >
> > The ML module will be stable since next release 2.8, also we have no
> > performance report to compare for example with Spark ML
> > Based on my exploration the performance of in terms of Big O notation is
> > the same like in Spark ML (real numbers says that Ignite ML is more
> faster
> > than Spark ML due to Ignite in-memory nature and so on)
> >
> > Since 2.8 it will have good integration with TensorFlow, Spark ML,
> XGBoost
> > via model inference.
> >
> > You as a user have no ability to run, for-example scikit-learn or R
> > packages in distributed mode over Ignite, but you could run the
> TensorFlow,
> > using Ignite as a distributed back-end instead of Horovod.
> >
> > If you have any questions, please let me know
> >
> >
> >
> > пт, 13 сент. 2019 г. в 21:28, Denis Magda :
> >
> >> David,
> >>
> >> Let me loop in Ignite dev list that has Ignite ML experts subscribed.
> >> Please, could you share more details in regards to your performance
> >> testing
> >> and objectives for Ignite ML overall?
> >>
> >> The module is ready for production and we're ready to help solve any
> >> cornerstones.
> >>
> >> -
> >> Denis
> >>
> >>
> >> On Fri, Sep 6, 2019 at 4:50 AM David Williams 
> >> wrote:
> >>
> >> > Python is 25 times slower than Java for ML at runtimes, which I found
> >> out
> >> > online. But I don't know that statement is true or not. I need
> insiders'
> >> > opinion.  Which ml other packages are best options for Ignite?
> >> >
> >> > On Fri, Sep 6, 2019 at 7:28 PM Mikael 
> >> wrote:
> >> >
> >> >> Hi!
> >> >>
> >> >> I have never used it myself but it's been there for long time and I
> >> >> would expect it to be stable, and yes it will run distributed, I
> can't
> >> >> say anything about performance as I have never used it.
> >> >>
> >> >> You will find a lot of more information at:
> >> >>
> >> >> https://apacheignite.readme.io/docs/machine-learning
> >> >>
> >> >> Mikael
> >> >>
> >> >>
> >> >> Den 2019-09-06 kl. 11:50, skrev David Williams:
> >> >> >
> >> >> >
> >> >> > I am evaluating ML framework for Java platform. I knew Ignite has
> ML
> >> >> > package.
> >> >> > But I like to know its stability and performance for production.
> Can
> >> >> > Ignite
> >> >> > ML code run in distribute way?
> >> >> >
> >> >> > Except its own ML package, which ml packages are best options for
> >> >> Ignite?
> >> >>
> >> >
> >>
> >
>

Re: ML stable and performance

[Alexey Zinoviev]

I could answer as one of developers of ML module.
Currently is available the ML in version 2.7.5, it supports a lot of
algorithms and could be used in production, but the API is not stable and
will be changed in 2.8

The ML module will be stable since next release 2.8, also we have no
performance report to compare for example with Spark ML
Based on my exploration the performance of in terms of Big O notation is
the same like in Spark ML (real numbers says that Ignite ML is more faster
than Spark ML due to Ignite in-memory nature and so on)

Since 2.8 it will have good integration with TensorFlow, Spark ML, XGBoost
via model inference.

You as a user have no ability to run, for-example scikit-learn or R
packages in distributed mode over Ignite, but you could run the TensorFlow,
using Ignite as a distributed back-end instead of Horovod.

If you have any questions, please let me know



пт, 13 сент. 2019 г. в 21:28, Denis Magda :

> David,
>
> Let me loop in Ignite dev list that has Ignite ML experts subscribed.
> Please, could you share more details in regards to your performance testing
> and objectives for Ignite ML overall?
>
> The module is ready for production and we're ready to help solve any
> cornerstones.
>
> -
> Denis
>
>
> On Fri, Sep 6, 2019 at 4:50 AM David Williams  wrote:
>
> > Python is 25 times slower than Java for ML at runtimes, which I found out
> > online. But I don't know that statement is true or not. I need insiders'
> > opinion.  Which ml other packages are best options for Ignite?
> >
> > On Fri, Sep 6, 2019 at 7:28 PM Mikael  wrote:
> >
> >> Hi!
> >>
> >> I have never used it myself but it's been there for long time and I
> >> would expect it to be stable, and yes it will run distributed, I can't
> >> say anything about performance as I have never used it.
> >>
> >> You will find a lot of more information at:
> >>
> >> https://apacheignite.readme.io/docs/machine-learning
> >>
> >> Mikael
> >>
> >>
> >> Den 2019-09-06 kl. 11:50, skrev David Williams:
> >> >
> >> >
> >> > I am evaluating ML framework for Java platform. I knew Ignite has ML
> >> > package.
> >> > But I like to know its stability and performance for production. Can
> >> > Ignite
> >> > ML code run in distribute way?
> >> >
> >> > Except its own ML package, which ml packages are best options for
> >> Ignite?
> >>
> >
>