From predragp at andrew.cmu.edu  Wed Nov  1 15:38:54 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Wed, 01 Nov 2017 15:38:54 -0400
Subject: git.int.autonlab.org updated
Message-ID: <20171101193854.N8MO5BqF8%predragp@andrew.cmu.edu>

Dear Autonians,

Please remove git.int.autonlab.org from .ssh/known_hosts and clear the
cash of your browsers to start using upgraded version Gogs/git.  Please
report any odd behavior right away.  The old version is still up and
running but I am the only one who can see it. I will keep it a day or
two for safety reasons.

Best,
Predrag

From predragp at andrew.cmu.edu  Wed Nov  1 22:58:58 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Wed, 01 Nov 2017 22:58:58 -0400
Subject: git.int.autonlab.org updated
In-Reply-To: <CAECAi++SXTa6F=1bOyj6Y1YJCwOyg7OUt-KGMFOtYKsVQrgEyQ@mail.gmail.com>
References: <20171101193854.N8MO5BqF8%predragp@andrew.cmu.edu>
 <CAECAi++SXTa6F=1bOyj6Y1YJCwOyg7OUt-KGMFOtYKsVQrgEyQ@mail.gmail.com>
Message-ID: <20171102025858.uy4nqNNfl%predragp@andrew.cmu.edu>

Donghan Wang <donghanw at cs.cmu.edu> wrote:

> Hi Predrag,
> 
> It seems bash is not available on the git server.
> 
> The issue occurs when you push a local branch to a remote repository.  To
> reproduce the issue, run the following commands
> 
> On your workstation, check out any repository
> 
> > git clone ssh://git at git.int.autonlab.org:2222/MCBA/ui.git
> 
> # create a new local branch "new-branch"
> > git checkout -b new-branch
> 
> # push it to the remote repo
> > git push origin new-branch
> 
> You will see error messages like this
> 
> Total 0 (delta 0), reused 0 (delta 0)
> remote: env: bash: No such file or directory
> To ssh://git at git.int.autonlab.org:2222/MCBA/ui.git
>  ! [remote rejected] new-branch -> new-branch (pre-receive hook declined)
> 

I just changed the default shell of user git to bash and restarted the
Gogs server. Please try again and see if you can reproduce the problem.

Cheers,
Predrag


> Thanks,
> Jarod
> 
> On Wed, Nov 1, 2017 at 3:38 PM, Predrag Punosevac <predragp at andrew.cmu.edu>
> wrote:
> 
> > Dear Autonians,
> >
> > Please remove git.int.autonlab.org from .ssh/known_hosts and clear the
> > cash of your browsers to start using upgraded version Gogs/git.  Please
> > report any odd behavior right away.  The old version is still up and
> > running but I am the only one who can see it. I will keep it a day or
> > two for safety reasons.
> >
> > Best,
> > Predrag
> >

From donghanw at cs.cmu.edu  Thu Nov  2 09:25:50 2017
From: donghanw at cs.cmu.edu (Donghan Wang)
Date: Thu, 2 Nov 2017 09:25:50 -0400
Subject: git.int.autonlab.org updated
In-Reply-To: <20171102025858.uy4nqNNfl%predragp@andrew.cmu.edu>
References: <20171101193854.N8MO5BqF8%predragp@andrew.cmu.edu>
 <CAECAi++SXTa6F=1bOyj6Y1YJCwOyg7OUt-KGMFOtYKsVQrgEyQ@mail.gmail.com>
 <20171102025858.uy4nqNNfl%predragp@andrew.cmu.edu>
Message-ID: <CAECAi+++uVCA3GdfR5=bBuFoTO2p1+KWzKnxpPPxgi=Fb0VmDQ@mail.gmail.com>

Hi Predrag,

The problem is resolved. I was able to push local branches to remote
repositories. Thank you!

Thanks,
Jarod

On Wed, Nov 1, 2017 at 10:58 PM, Predrag Punosevac <predragp at andrew.cmu.edu>
wrote:

> Donghan Wang <donghanw at cs.cmu.edu> wrote:
>
> > Hi Predrag,
> >
> > It seems bash is not available on the git server.
> >
> > The issue occurs when you push a local branch to a remote repository.  To
> > reproduce the issue, run the following commands
> >
> > On your workstation, check out any repository
> >
> > > git clone ssh://git at git.int.autonlab.org:2222/MCBA/ui.git
> >
> > # create a new local branch "new-branch"
> > > git checkout -b new-branch
> >
> > # push it to the remote repo
> > > git push origin new-branch
> >
> > You will see error messages like this
> >
> > Total 0 (delta 0), reused 0 (delta 0)
> > remote: env: bash: No such file or directory
> > To ssh://git at git.int.autonlab.org:2222/MCBA/ui.git
> >  ! [remote rejected] new-branch -> new-branch (pre-receive hook declined)
> >
>
> I just changed the default shell of user git to bash and restarted the
> Gogs server. Please try again and see if you can reproduce the problem.
>
> Cheers,
> Predrag
>
>
> > Thanks,
> > Jarod
> >
> > On Wed, Nov 1, 2017 at 3:38 PM, Predrag Punosevac <
> predragp at andrew.cmu.edu>
> > wrote:
> >
> > > Dear Autonians,
> > >
> > > Please remove git.int.autonlab.org from .ssh/known_hosts and clear the
> > > cash of your browsers to start using upgraded version Gogs/git.  Please
> > > report any odd behavior right away.  The old version is still up and
> > > running but I am the only one who can see it. I will keep it a day or
> > > two for safety reasons.
> > >
> > > Best,
> > > Predrag
> > >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171102/e2c8f21f/attachment.html>

From awd at cs.cmu.edu  Thu Nov  2 10:52:51 2017
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Thu, 2 Nov 2017 10:52:51 -0400
Subject: =?UTF-8?Q?Fwd:_You=e2=80=99re_Invited:_CMU_AI_All-Hands_Meeting_|_N?=
 =?UTF-8?Q?ov_2_at_1:00pm_|_McConomy_Auditorium?=
In-Reply-To: <CAOQLAA1oLYv77wRLpN_oXL1ia+eaPsWvFpHM+u2vNrBPDM9tKw@mail.gmail.com>
References: <CAOQLAA1oLYv77wRLpN_oXL1ia+eaPsWvFpHM+u2vNrBPDM9tKw@mail.gmail.com>
Message-ID: <82ce29ac-089f-70f7-49b7-4fe6ef1045b4@cs.cmu.edu>

Just a reminder about this cool event happening today.

Artur


-------- Forwarded Message --------
Subject: 	You?re Invited: CMU AI All-Hands Meeting | Nov 2 at 1:00pm | 
McConomy Auditorium
Date: 	Tue, 24 Oct 2017 17:11:28 -0400
From: 	Andrew Moore <awm at cs.cmu.edu>
To: 	scs-faculty at cs.cmu.edu



Dear SCS Faculty,

Please join us next Thursday?for the official campus launch of the CMU 
AI initiative! It's going to be fun and will provide an opportunity for 
everyone doing AI across campus (including these folks 
<http://ai.cs.cmu.edu/faculty-listing>) to get to know about each other 
and strengthen interactions.

Please forward this invitation to other folks who you think would like 
to attend, including your students, postdocs and staff. If you notice 
we're missing folks from the AI Faculty Listing (link above), please let 
Nichole Merritt <mailto:ncmerritt at cmu.edu>?know!

Thanks,
Andrew
/*You?re Invited! */

*CMU AI All-Hands Meeting*
*Thursday, November 2, 2017*
*1:00-2:30pm**
*McConomy Auditorium, Cohon University Center*
/?Reception to follow in Rangos 3/
/
/

Join us for the official campus launch of the CMU AI initiative on 
November 2nd at 1:00pm?in McConomy Auditorium!

The program will include a reminder of the heart and soul of AI at CMU. 
We will develop a roadmap of what is happening with AI on scientific, 
academic and societal fronts, discuss AI education, and present some of 
CMU's most extraordinary AI efforts. Through panels, examples and 
audience participation we will identify some of the best areas to 
potentially create AI ?moonshots,? and we will attempt to start framing 
a campus-wide
"matching service" in which students, faculty and staff can find 
compatible collaborators in unexpected, new areas across campus.

Planning to attend?**
RSVP here! <https://goo.gl/UKtD2Y>

Learn more before the event. Visit us at
*ai.cs.cmu.edu <http://ai.cs.cmu.edu/>*
*
*
Sincerely,
Jaime Carbonell ?| ?Martial Hebert ?| ?Andrew Moore ?| ?Tuomas Sandholm 
 ?| ?Manuela Veloso
*
*

-- 
Andrew W. Moore <http://www.cs.cmu.edu/%7Eawm/> | Professor and Dean, 
School of Computer Science <http://www.cs.cmu.edu/> | Carnegie Mellon 
University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171102/171ca3d0/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cmu.AI-EmailBanner_171002-02[5][1].png
Type: image/png
Size: 195992 bytes
Desc: not available
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171102/171ca3d0/attachment-0001.png>

From yhechtli at andrew.cmu.edu  Mon Nov 13 12:37:35 2017
From: yhechtli at andrew.cmu.edu (Yotam Hechtlinger)
Date: Mon, 13 Nov 2017 12:37:35 -0500
Subject: Set up local python virtual environment
Message-ID: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>

Hello Everyone,

I'm a new user to the Autonlab, right now setting up a working environment.
Predrag has told some user might have scripts installing most of the
relevant python packages to do neural network computations.

I wonder if anyone can share one of those?
It will greatly save me configuration time.

Thanks a lot,
Yotam.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171113/a91eb9ce/attachment.html>

From mbarnes1 at andrew.cmu.edu  Mon Nov 13 12:52:25 2017
From: mbarnes1 at andrew.cmu.edu (Matthew Barnes)
Date: Mon, 13 Nov 2017 17:52:25 +0000
Subject: Set up local python virtual environment
In-Reply-To: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
References: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
Message-ID: <CAB7OVwCoPS1cL+SiO3BGHQHH8kdbJF2SWqJKfogGNt5yN5xCUA@mail.gmail.com>

Hey Yotam,
I'd highly recommend using Conda, which will manage both your virtual
environments and packages. Predrag has Conda installed on all the GPU
machines (and maybe the CPU machines).

Once you've figured out how to use Conda, installing all the packages you
need for a particular deep learning framework (e.g. Tensorflow, PyTorch) is
a single command.

- Matt

On Mon, Nov 13, 2017 at 12:39 PM Yotam Hechtlinger <yhechtli at andrew.cmu.edu>
wrote:

> Hello Everyone,
>
> I'm a new user to the Autonlab, right now setting up a working
> environment.
> Predrag has told some user might have scripts installing most of the
> relevant python packages to do neural network computations.
>
> I wonder if anyone can share one of those?
> It will greatly save me configuration time.
>
> Thanks a lot,
> Yotam.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171113/d20426c3/attachment.html>

From sheath at andrew.cmu.edu  Mon Nov 13 12:57:06 2017
From: sheath at andrew.cmu.edu (Simon Heath)
Date: Mon, 13 Nov 2017 12:57:06 -0500
Subject: Set up local python virtual environment
In-Reply-To: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
References: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
Message-ID: <CAGeWgZTjdo1c61XpR4YTzBQMZ8P=SR=aKK3Tyk_gR2n9hMFRrQ@mail.gmail.com>

Hi Yotam,

The tool you want to be using is `virtualenv`, which should be installed on
all the compute nodes.  Documentation is here:
https://virtualenv.pypa.io/en/stable/

It allows you to easily install whatever python packages you need in a
subdirectory, so you can install things without needing root permissions.
I recommend just making a shell script for each project to create a
virtualenv with any packages you need to install.

Simon

On Mon, Nov 13, 2017 at 12:37 PM, Yotam Hechtlinger <yhechtli at andrew.cmu.edu
> wrote:

> Hello Everyone,
>
> I'm a new user to the Autonlab, right now setting up a working
> environment.
> Predrag has told some user might have scripts installing most of the
> relevant python packages to do neural network computations.
>
> I wonder if anyone can share one of those?
> It will greatly save me configuration time.
>
> Thanks a lot,
> Yotam.
>
>


-- 
Simon Heath, Research Programmer and Analyst
Robotics Institute - Auton Lab
Carnegie Mellon University
sheath at andrew.cmu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171113/f2413018/attachment.html>

From sheath at andrew.cmu.edu  Mon Nov 13 12:58:35 2017
From: sheath at andrew.cmu.edu (Simon Heath)
Date: Mon, 13 Nov 2017 12:58:35 -0500
Subject: Set up local python virtual environment
In-Reply-To: <CAB7OVwCoPS1cL+SiO3BGHQHH8kdbJF2SWqJKfogGNt5yN5xCUA@mail.gmail.com>
References: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
 <CAB7OVwCoPS1cL+SiO3BGHQHH8kdbJF2SWqJKfogGNt5yN5xCUA@mail.gmail.com>
Message-ID: <CAGeWgZRDr-fUbXfVvJZbXf+ePWQfsxr4wiL8+f-X+1u=FgqqLg@mail.gmail.com>

Okay, Conda is a good option too; I just don't know it well.

Simon

On Mon, Nov 13, 2017 at 12:52 PM, Matthew Barnes <mbarnes1 at andrew.cmu.edu>
wrote:

> Hey Yotam,
> I'd highly recommend using Conda, which will manage both your virtual
> environments and packages. Predrag has Conda installed on all the GPU
> machines (and maybe the CPU machines).
>
> Once you've figured out how to use Conda, installing all the packages you
> need for a particular deep learning framework (e.g. Tensorflow, PyTorch) is
> a single command.
>
> - Matt
>
> On Mon, Nov 13, 2017 at 12:39 PM Yotam Hechtlinger <
> yhechtli at andrew.cmu.edu> wrote:
>
>> Hello Everyone,
>>
>> I'm a new user to the Autonlab, right now setting up a working
>> environment.
>> Predrag has told some user might have scripts installing most of the
>> relevant python packages to do neural network computations.
>>
>> I wonder if anyone can share one of those?
>> It will greatly save me configuration time.
>>
>> Thanks a lot,
>> Yotam.
>>
>>


-- 
Simon Heath, Research Programmer and Analyst
Robotics Institute - Auton Lab
Carnegie Mellon University
sheath at andrew.cmu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171113/7b65f5ae/attachment-0001.html>

From mbarnes1 at andrew.cmu.edu  Mon Nov 13 13:30:23 2017
From: mbarnes1 at andrew.cmu.edu (Matthew Barnes)
Date: Mon, 13 Nov 2017 18:30:23 +0000
Subject: Set up local python virtual environment
In-Reply-To: <CAGeWgZTjdo1c61XpR4YTzBQMZ8P=SR=aKK3Tyk_gR2n9hMFRrQ@mail.gmail.com>
References: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
 <CAGeWgZTjdo1c61XpR4YTzBQMZ8P=SR=aKK3Tyk_gR2n9hMFRrQ@mail.gmail.com>
Message-ID: <CAB7OVwAE1bz2NSavUm9hubaVt+EyLxmphEyjt4k2MQ6CSLVwWw@mail.gmail.com>

Virtualenv is more classic,  At least for PyTorch, the Conda install is way
easier. I believe Tensorflow is also simpler with Conda.

On Mon, Nov 13, 2017 at 12:58 PM Simon Heath <sheath at andrew.cmu.edu> wrote:

> Hi Yotam,
>
> The tool you want to be using is `virtualenv`, which should be installed
> on all the compute nodes.  Documentation is here:
> https://virtualenv.pypa.io/en/stable/
>
> It allows you to easily install whatever python packages you need in a
> subdirectory, so you can install things without needing root permissions.
> I recommend just making a shell script for each project to create a
> virtualenv with any packages you need to install.
>
> Simon
>
> On Mon, Nov 13, 2017 at 12:37 PM, Yotam Hechtlinger <
> yhechtli at andrew.cmu.edu> wrote:
>
>> Hello Everyone,
>>
>> I'm a new user to the Autonlab, right now setting up a working
>> environment.
>> Predrag has told some user might have scripts installing most of the
>> relevant python packages to do neural network computations.
>>
>> I wonder if anyone can share one of those?
>> It will greatly save me configuration time.
>>
>> Thanks a lot,
>> Yotam.
>>
>>
>
>
> --
> Simon Heath, Research Programmer and Analyst
> Robotics Institute - Auton Lab
> Carnegie Mellon University
> sheath at andrew.cmu.edu
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171113/c2692a11/attachment.html>

From predragp at andrew.cmu.edu  Mon Nov 13 14:13:02 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Mon, 13 Nov 2017 14:13:02 -0500
Subject: Set up local python virtual environment
In-Reply-To: <CAB7OVwCoPS1cL+SiO3BGHQHH8kdbJF2SWqJKfogGNt5yN5xCUA@mail.gmail.com>
References: <CABftdoD0SK64Kq-sfURx8XYV8jgoo2cXWgbCDh=AqH6GZb5EUA@mail.gmail.com>
 <CAB7OVwCoPS1cL+SiO3BGHQHH8kdbJF2SWqJKfogGNt5yN5xCUA@mail.gmail.com>
Message-ID: <20171113191302.u2cwMc4lt%predragp@andrew.cmu.edu>

Matthew Barnes <mbarnes1 at andrew.cmu.edu> wrote:

> Hey Yotam,
> I'd highly recommend using Conda, which will manage both your virtual
> environments and packages. Predrag has Conda installed on all the GPU
> machines (and maybe the CPU machines).
> 
> Once you've figured out how to use Conda, installing all the packages you
> need for a particular deep learning framework (e.g. Tensorflow, PyTorch) is
> a single command.
> 
> - Matt

Mini Conda has being creating problems. IIRC I left it only on one the
servers (GPU1 or GPU2 I no longer remember) for you. Please try to use
virtualenv from /opt/rh/devtools4 if you need python 3.5 or to use just
a regular virtualenv which is the part of OS for 2.7 branch.

Predrag


> 
> On Mon, Nov 13, 2017 at 12:39 PM Yotam Hechtlinger <yhechtli at andrew.cmu.edu>
> wrote:
> 
> > Hello Everyone,
> >
> > I'm a new user to the Autonlab, right now setting up a working
> > environment.
> > Predrag has told some user might have scripts installing most of the
> > relevant python packages to do neural network computations.
> >
> > I wonder if anyone can share one of those?
> > It will greatly save me configuration time.
> >
> > Thanks a lot,
> > Yotam.
> >
> >

From awd at cs.cmu.edu  Thu Nov 16 14:29:27 2017
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Thu, 16 Nov 2017 14:29:27 -0500
Subject: Fwd: [Jmlr-announce] Classification of Time Sequences using
 Graphs of Temporal Constraints
In-Reply-To: <CAOQ-VV2=9qK+SL2w1v52JeJS1vK9w1dm-eBiXREgKFsz7w5VvA@mail.gmail.com>
References: <201711161811.vAGIBsbA012084@lmtp1.ucs.ed.ac.uk>
 <CAOQ-VV2=9qK+SL2w1v52JeJS1vK9w1dm-eBiXREgKFsz7w5VvA@mail.gmail.com>
Message-ID: <88d716ec-f86b-5b57-2967-3e97b72ad475@cs.cmu.edu>

This is probably the longest-reviewed paper ever submitted by an Autonian!
It feels like we were submitting it sometime a century ago :P

But the ever-elusive GTC paper is finally? out, and I am very proud of 
Mathieu
who persevered through this challenging experience!

Cheers,
Artur

PS If this paper is setting the record for the amount of time spent in 
the editor's hands,
let us keep it this way and avoid challenging it with another submission :)

On 11/16/2017 2:15 PM, Lujie Chen wrote:
> Finally saw it in press!
> ---------- Forwarded message ----------
> From: *Charles Sutton* <production at jmlr.org <mailto:production at jmlr.org>>
> Date: Thu, Nov 16, 2017 at 1:11 PM
> Subject: [Jmlr-announce] Classification of Time Sequences using Graphs 
> of Temporal Constraints
> To: jmlr-announce at csail.mit.edu <mailto:jmlr-announce at csail.mit.edu>
> Cc: production at jmlr.org <mailto:production at jmlr.org>
>
>
> The Journal of Machine Learning Research (www.jmlr.org 
> <http://www.jmlr.org>) is pleased to announce the publication of a new 
> paper:
> ------------------------------------------------------------------------------
> Classification of Time Sequences using Graphs of Temporal Constraints
> Mathieu Guillame-Bert, Artur Dubrawski
> JMLR 18(121):1-34, 2017.
> Link: http://jmlr.org/papers/v18/15-403.html 
> <http://jmlr.org/papers/v18/15-403.html>
>
> Abstract
> We introduce two algorithms that learn to classify Symbolic and Scalar 
> Time Sequences (SSTS); an extension of multivariate time series. An 
> SSTS is a set of \emph{events} and a set of <em>scalars</em>. An 
> <em>event</em> is defined by a symbol and a time-stamp. A 
> <em>scalar</em> is defined by a symbol and a function mapping a number 
> for each possible time stamp of the data. The proposed algorithms rely 
> on temporal patterns called Graph of Temporal Constraints (GTC). A GTC 
> is a directed graph in which vertices express occurrences of specific 
> events, and edges express temporal constraints between occurrences of 
> pairs of events. Additionally, each vertex of a GTC can be augmented 
> with numeric constraints on scalar values. We allow GTCs to be cyclic 
> and/or disconnected. The first of the introduced algorithms extracts 
> sets of co-dependent GTCs to be used in a voting mechanism. The second 
> algorithm builds decision forest like representations where each node 
> is a GTC. In both algorit!
> ?hms, extraction of GTCs and model building are interleaved. Both 
> algorithms are closely related to each other and they exhibit 
> complementary properties including complexity, performance, and 
> interpretability. The main novelties of this work reside in direct 
> building of the model and efficient learning of GTC structures. We 
> explain the proposed algorithms and evaluate their performance against 
> a diverse collection of 59 benchmark data sets. In these experiments, 
> our algorithms come across as highly competitive and in most cases 
> closely match or outperform state-of-the-art alternatives in terms of 
> the computational speed while dominating in terms of the accuracy of 
> classification of time sequences.
>
> ------------------------------------------------------------------------------
> This paper and previous papers are available electronically at 
> http://www.jmlr.org in PDF format. The papers of Volumes 1-4 were also 
> published in hardcopy by MIT Press; please see 
> http://mitpress.mit.edu/JMLR for details. Volume 5 and subsequent 
> volumes are being printed in hardcopy by Microtome Publishing. Please 
> see http://www.mtome.com/Publications/JMLR/jmlr.html 
> <http://www.mtome.com/Publications/JMLR/jmlr.html> for details and 
> ordering information.
>
> Charles Sutton
> production at jmlr.org <mailto:production at jmlr.org>
>
> _______________________________________________
> Jmlr-announce mailing list
> Jmlr-announce at lists.csail.mit.edu 
> <mailto:Jmlr-announce at lists.csail.mit.edu>
> https://lists.csail.mit.edu/mailman/listinfo/jmlr-announce 
> <https://lists.csail.mit.edu/mailman/listinfo/jmlr-announce>
>
>
>
> -- 
>
> ==================
>
> Karen (Lujie) Chen
>
> Ph.D. Candidate in Information Systems, Heinz College
> PIER (Program of Interdisciplinary Educational Research)
> Member of Auton Lab, Robotics Institute
> Newell-Simon Hall 3124
> Carnegie Mellon University
> Pittsburgh, PA 15213
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171116/e67177a0/attachment.html>

From awd at cs.cmu.edu  Tue Nov 21 11:56:29 2017
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Tue, 21 Nov 2017 11:56:29 -0500
Subject: Fwd: RI Ph.D. Thesis Proposal: Matt Barnes
In-Reply-To: <23856_1511192278_vAKFbvFl023875_1511192238294.44077@cmu.edu>
References: <23856_1511192278_vAKFbvFl023875_1511192238294.44077@cmu.edu>
Message-ID: <d393a7cd-aa11-0cf1-2d84-89a9e863760c@cs.cmu.edu>

Team,

Happy Thanksgiving!

+ please mark your calendars for Monday next week.
Attending Matt's proposal talk will surely help us burn the excess 
calories from our turkey dinners.

Cheers,
Artur


-------- Forwarded Message --------
Subject: 	RI Ph.D. Thesis Proposal: Matt Barnes
Date: 	Mon, 20 Nov 2017 15:37:18 +0000
From: 	Suzanne Muth <lyonsmuth at cmu.edu>
To: 	ri-people at cs.cmu.edu <ri-people at cs.cmu.edu>



Date: ??27?November 2017

Time: ? 4:00 p.m.

Place: ?Gates Hillman Center 4405

Type: ??Ph.D. Thesis Proposal

Who: ??Matt Barnes

Topic: ?Learning with Clusters


Abstract:

As machine learning becomes more ubiquitous, clustering has evolved from 
primarily a data analysis tool into an integrated component of complex 
machine learning systems, including those involving dimensionality 
reduction, anomaly detection, network analysis, image segmentation and 
classifying groups of data. With this integration into multi-stage 
systems comes a need to better understand interactions between pipeline 
components. Changing parameters of the clustering algorithm will impact 
downstream components and, quite unfortunately, it is usually not 
possible to simply back-propagate through the entire system. Currently, 
as with many machine learning systems, the output of the clustering 
algorithm is taken as ground truth at the next pipeline step. Our 
empirical results show this false assumption may have dramatic empirical 
impacts -- sometimes biasing results by upwards of 25%.

We address this gap by developing estimators and methods to both 
quantify and correct for clustering errors' impacts on downstream 
learners. Our work is agnostic to the downstream learners, and requires 
few assumptions on the clustering algorithm. Theoretical and empirical 
results demonstrate our methods and estimators are superior to the 
current naive approaches, which do not account for clustering errors.
?
Along these lines, we also develop several new clustering algorithms and 
prove theoretical bounds for existing algorithms, to be used as inputs 
to our later error-correction methods. Not surprisingly, we find 
learning on clusters of data is both theoretically and empirically 
easier as the number of clustering errors decreases. Thus, our work is 
two-fold. We attempt to both provide the best clustering possible and 
learn on inevitably noisy clusters.

A major limiting factor in our error-correction methods is scalability. 
Currently, their computational complexity is O(n^3) where n is the size 
of the training dataset. This limits their applicability to very small 
machine learning problems. We propose addressing this scalability issue 
through approximation. It should be possible to reduce the computational 
complexity to O(p^3) where p is a small fixed constant and independent 
of n, corresponding to the number of parameters in the approximation.



Thesis Committee Members:

Artur Dubrawski, Chair

Geoff Gordon

Kris Kitani

Beka Steorts, Duke University



A copy of the thesis proposal document is available at:

http://goo.gl/MpwTCN

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171121/c6259658/attachment.html>

From mbarnes1 at andrew.cmu.edu  Sun Nov 26 13:48:16 2017
From: mbarnes1 at andrew.cmu.edu (Matthew Barnes)
Date: Sun, 26 Nov 2017 18:48:16 +0000
Subject: Presentation clicker
Message-ID: <CAB7OVwAK2B87rwq--a_WDGu4tsD-aAysdTi3j7b9L4BKvDQ91A@mail.gmail.com>

Anyone have one of those remote presentation clickers? I'd like to borrow
it for tomorrow.

Thanks,
Matt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171126/33429604/attachment.html>

From awd at cs.cmu.edu  Mon Nov 27 10:36:52 2017
From: awd at cs.cmu.edu (Artur Dubrawski)
Date: Mon, 27 Nov 2017 10:36:52 -0500
Subject: Fwd: RI Ph.D. Thesis Proposal: Matt Barnes
In-Reply-To: <d393a7cd-aa11-0cf1-2d84-89a9e863760c@cs.cmu.edu>
References: <23856_1511192278_vAKFbvFl023875_1511192238294.44077@cmu.edu>
 <d393a7cd-aa11-0cf1-2d84-89a9e863760c@cs.cmu.edu>
Message-ID: <f0fd0854-278e-7209-a3dd-00912b610135@cs.cmu.edu>

It is today, 4pm, Gates 4405!

A.

On 11/21/2017 11:56 AM, Artur Dubrawski wrote:
> Team,
>
> Happy Thanksgiving!
>
> + please mark your calendars for Monday next week.
> Attending Matt's proposal talk will surely help us burn the excess 
> calories from our turkey dinners.
>
> Cheers,
> Artur
>
>
> -------- Forwarded Message --------
> Subject: 	RI Ph.D. Thesis Proposal: Matt Barnes
> Date: 	Mon, 20 Nov 2017 15:37:18 +0000
> From: 	Suzanne Muth <lyonsmuth at cmu.edu>
> To: 	ri-people at cs.cmu.edu <ri-people at cs.cmu.edu>
>
>
>
> Date: ??27?November 2017
>
> Time: 4:00 p.m.
>
> Place: ?Gates Hillman Center 4405
>
> Type: ??Ph.D. Thesis Proposal
>
> Who: ??Matt Barnes
>
> Topic: ?Learning with Clusters
>
>
> Abstract:
>
> As machine learning becomes more ubiquitous, clustering has evolved 
> from primarily a data analysis tool into an integrated component of 
> complex machine learning systems, including those involving 
> dimensionality reduction, anomaly detection, network analysis, image 
> segmentation and classifying groups of data. With this integration 
> into multi-stage systems comes a need to better understand 
> interactions between pipeline components. Changing parameters of the 
> clustering algorithm will impact downstream components and, quite 
> unfortunately, it is usually not possible to simply back-propagate 
> through the entire system. Currently, as with many machine learning 
> systems, the output of the clustering algorithm is taken as ground 
> truth at the next pipeline step. Our empirical results show this false 
> assumption may have dramatic empirical impacts -- sometimes biasing 
> results by upwards of 25%.
>
> We address this gap by developing estimators and methods to both 
> quantify and correct for clustering errors' impacts on downstream 
> learners. Our work is agnostic to the downstream learners, and 
> requires few assumptions on the clustering algorithm. Theoretical and 
> empirical results demonstrate our methods and estimators are superior 
> to the current naive approaches, which do not account for clustering 
> errors.
> ?
> Along these lines, we also develop several new clustering algorithms 
> and prove theoretical bounds for existing algorithms, to be used as 
> inputs to our later error-correction methods. Not surprisingly, we 
> find learning on clusters of data is both theoretically and 
> empirically easier as the number of clustering errors decreases. Thus, 
> our work is two-fold. We attempt to both provide the best clustering 
> possible and learn on inevitably noisy clusters.
>
> A major limiting factor in our error-correction methods is 
> scalability. Currently, their computational complexity is O(n^3) where 
> n is the size of the training dataset. This limits their applicability 
> to very small machine learning problems. We propose addressing this 
> scalability issue through approximation. It should be possible to 
> reduce the computational complexity to O(p^3) where p is a small fixed 
> constant and independent of n, corresponding to the number of 
> parameters in the approximation.
>
>
>
> Thesis Committee Members:
>
> Artur Dubrawski, Chair
>
> Geoff Gordon
>
> Kris Kitani
>
> Beka Steorts, Duke University
>
>
>
> A copy of the thesis proposal document is available at:
>
> http://goo.gl/MpwTCN
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171127/1471bf3a/attachment.html>

From sheath at andrew.cmu.edu  Thu Nov 30 11:18:34 2017
From: sheath at andrew.cmu.edu (Simon Heath)
Date: Thu, 30 Nov 2017 11:18:34 -0500
Subject: git 2.9 installed on compute nodes
Message-ID: <CAGeWgZQPGdYiEc1S5h6zYXq9KG0HXiSUgUsGSDUp4zr2kjm8vA@mail.gmail.com>

Hi all,

git 2.9 should now be installed on all compute nodes.  It can be found at
/opt/rh/rh-git29/root/usr/bin/git .  If you have software or scripts that
need newer versions of git than the system default (1.8), you can tell them
to use this one.

Simon

-- 
Simon Heath, Research Programmer and Analyst
Robotics Institute - Auton Lab
Carnegie Mellon University
sheath at andrew.cmu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171130/50e7ac94/attachment.html>

From predragp at andrew.cmu.edu  Thu Nov 30 17:45:41 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Thu, 30 Nov 2017 17:45:41 -0500
Subject: nvidia-smi not available on gpu1, gpu5-7
In-Reply-To: <CFEF5945-158F-4D3F-875F-73CFA558AA52@cs.cmu.edu>
References: <CFEF5945-158F-4D3F-875F-73CFA558AA52@cs.cmu.edu>
Message-ID: <20171130224541.G_5PxX8B5%predragp@andrew.cmu.edu>

Yichong Xu <yichongx at cs.cmu.edu> wrote:

> Hi Predrag,
> It seems nvidia-smi is not running properly on gpu1 and gpu5-7. (My programs are still running but I???m not sure whether they???re still using the gpu or not.) Could you check about this? Thank you very much!
> 

Thanks for the report. Apparently CUDA upgrade broke the something

root at gpu7$ nvidia-smi
Failed to initialize NVML: Driver/library version mismatch

I am working to fix the issue.

Predrag

> Yichong Xu
> Machine Learning Department, CMU
> yichongx at cs.cmu.edu
> 412-652-8309
> 
> 
> 
> 

From yichongx at cs.cmu.edu  Thu Nov 30 18:04:21 2017
From: yichongx at cs.cmu.edu (Yichong Xu)
Date: Thu, 30 Nov 2017 18:04:21 -0500
Subject: nvidia-smi not available on gpu1, gpu5-7
In-Reply-To: <20171130224541.G_5PxX8B5%predragp@andrew.cmu.edu>
References: <CFEF5945-158F-4D3F-875F-73CFA558AA52@cs.cmu.edu>
 <20171130224541.G_5PxX8B5%predragp@andrew.cmu.edu>
Message-ID: <11248D72-FAE5-44B3-8D2E-666E3C1644C1@cs.cmu.edu>

Thank you Predrag!
Thanks,
Yichong



> On Nov 30, 2017, at 5:45 PM, Predrag Punosevac <predragp at andrew.cmu.edu> wrote:
> 
> Yichong Xu <yichongx at cs.cmu.edu> wrote:
> 
>> Hi Predrag,
>> It seems nvidia-smi is not running properly on gpu1 and gpu5-7. (My programs are still running but I???m not sure whether they???re still using the gpu or not.) Could you check about this? Thank you very much!
>> 
> 
> Thanks for the report. Apparently CUDA upgrade broke the something
> 
> root at gpu7$ nvidia-smi
> Failed to initialize NVML: Driver/library version mismatch
> 
> I am working to fix the issue.
> 
> Predrag
> 
>> Yichong Xu
>> Machine Learning Department, CMU
>> yichongx at cs.cmu.edu
>> 412-652-8309
>> 
>> 
>> 
>> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/autonlab-users/attachments/20171130/7791e255/attachment.html>

From predragp at andrew.cmu.edu  Thu Nov 30 18:19:04 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Thu, 30 Nov 2017 18:19:04 -0500
Subject: nvidia-smi not available on gpu1, gpu5-7
In-Reply-To: <11248D72-FAE5-44B3-8D2E-666E3C1644C1@cs.cmu.edu>
References: <CFEF5945-158F-4D3F-875F-73CFA558AA52@cs.cmu.edu>
 <20171130224541.G_5PxX8B5%predragp@andrew.cmu.edu>
 <11248D72-FAE5-44B3-8D2E-666E3C1644C1@cs.cmu.edu>
Message-ID: <20171130231904.Aul0YRwg2%predragp@andrew.cmu.edu>

Yichong Xu <yichongx at cs.cmu.edu> wrote:

> Thank you Predrag!
> Thanks,
> Yichong
> 

Dear Autonians,

I know how to fix this. Upgrade process have upgraded CUDA to 9.0 which
requires new 

NVIDIA-Linux-x86_64-384.98

driver instead of 

NVIDIA-Linux-x86_64-384.90

GPU7 now works as expected (I tested MATLAB). Be mindful that some other
peaces of custom compiled software (TensorFlow, Caffe) might have gotten
broken and need to be recompiled. 

I am upgrading driver right now on GPU5 and GPU6


GPU1 works also as exected (Tesla K80 cards) but CUDA is upgraded to 9.0
so crazy things might happen until dust settles.

Best,
Predrag

P.S. CUDA-9 is supposedly much faster than CUDA-8




> 
> 
> > On Nov 30, 2017, at 5:45 PM, Predrag Punosevac <predragp at andrew.cmu.edu> wrote:
> > 
> > Yichong Xu <yichongx at cs.cmu.edu> wrote:
> > 
> >> Hi Predrag,
> >> It seems nvidia-smi is not running properly on gpu1 and gpu5-7. (My programs are still running but I???m not sure whether they???re still using the gpu or not.) Could you check about this? Thank you very much!
> >> 
> > 
> > Thanks for the report. Apparently CUDA upgrade broke the something
> > 
> > root at gpu7$ nvidia-smi
> > Failed to initialize NVML: Driver/library version mismatch
> > 
> > I am working to fix the issue.
> > 
> > Predrag
> > 
> >> Yichong Xu
> >> Machine Learning Department, CMU
> >> yichongx at cs.cmu.edu
> >> 412-652-8309
> >> 
> >> 
> >> 
> >> 
> 

From predragp at andrew.cmu.edu  Thu Nov 30 21:53:26 2017
From: predragp at andrew.cmu.edu (Predrag Punosevac)
Date: Thu, 30 Nov 2017 21:53:26 -0500
Subject: Openbox available on x2go
Message-ID: <20171201025326.n4L7fL1ql%predragp@andrew.cmu.edu>

Dear Autonians,

For the past couple of months there were numerous complains about
usability of X2Go client caused by icewm bugs. I have installed Openbox
on all computing nodes and used extensively over the Thanksgiving break.
Openbox is super simple Window Manager for X server and fairly easy to
use so please give it a try if you have issues with icewm.

Best,
Predrag