verne.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
MystikNinja has joined #mlpack
MystikNinja has quit [Ping timeout: 276 seconds]
MystikNinja has joined #mlpack
< MystikNinja>
Are there any APIs in mlpack to generate data? How would I go about developing test cases to test convergence of the MVU+LRSDP functions?
< MystikNinja>
My first guess would be to simply prepare a Swiss roll type dataset and see visually what MVU returns.
thepro has joined #mlpack
< MystikNinja>
As in, look at some visualization of the output data and see if it has unrolled it "well enough"
< thepro>
hey, I am interested in the Gradient Tree Boosting and XGBoost algorithms. Are they good enough for a GSoC proposal?
Devon has joined #mlpack
< Devon>
Hello, I am interested in talking with a Google Summer of Code Mentor.
< Devon>
please contact me at fulcherd@southwestern.edu
Devon has quit [Client Quit]
MystikNinja has quit [Remote host closed the connection]
csoni has quit [Ping timeout: 264 seconds]
csoni has joined #mlpack
fsub has joined #mlpack
csoni has quit [Ping timeout: 260 seconds]
Asus_ has joined #mlpack
csoni has joined #mlpack
fsub has quit [Ping timeout: 260 seconds]
pktiw has joined #mlpack
Prabhat-IIT has joined #mlpack
teor4ad has joined #mlpack
teor4ad has quit [Client Quit]
< Prabhat-IIT>
zoq: As soon as you get some time please look into my proposal on PSO. I've addresed your valuable comments
pktiw has quit [Quit: Page closed]
prem_ktiw has joined #mlpack
csoni has quit [Ping timeout: 264 seconds]
csoni has joined #mlpack
csoni2 has joined #mlpack
csoni has quit [Read error: Connection reset by peer]
prem_ktiw has quit [Ping timeout: 240 seconds]
prem_ktiw has joined #mlpack
csoni2 has quit [Ping timeout: 240 seconds]
csoni has joined #mlpack
prem_ktiw has quit [Ping timeout: 246 seconds]
prem_ktiw has joined #mlpack
csoni has quit [Read error: Connection reset by peer]
csoni has joined #mlpack
csoni has quit [Read error: Connection reset by peer]
petris has quit [Read error: Connection reset by peer]
petris has joined #mlpack
petris has quit [Changing host]
petris has joined #mlpack
Prabhat-IIT has quit [Ping timeout: 260 seconds]
csoni2 has joined #mlpack
csoni has quit [Read error: Connection reset by peer]
csoni2 has quit [Ping timeout: 240 seconds]
Asus_ has quit [Ping timeout: 260 seconds]
csoni has joined #mlpack
thepro has quit [Quit: Leaving]
< Atharva>
In ffn implementation, in the train method, there is a line Log::info which prints the final objective, but when I trained a model, nothing was printed.
< Atharva>
I modified the code and added std::cout to print the objective, it was printed this time.
fsub has joined #mlpack
< Atharva>
Another thing, when we use the train function, at what point does it stop, how many iterations? Or some value of objective?
< Atharva>
Okay, I am clear about the 2nd doubt, there’s a default maxIterations value in the optimizer object.
sourabhvarshney1 has joined #mlpack
fsub has quit [Quit: Leaving]
sourabhvarshney1 has quit [Quit: Page closed]
satyam_2401 has joined #mlpack
< csoni>
Atharva: you need to pass in the -verbose flag when you run the program. Log::Info doesn't print to the screen by default.
< csoni>
Feel free to look up the I/O documentation online for more information.
csoni has quit [Read error: Connection reset by peer]
< yashsharan>
Hello everyone.I was thinking of writing my GSoc Proposal related to reinforcement learning.Should I get the proposal reviewed by someone before final submission?TIA.
vivekp has quit [Ping timeout: 240 seconds]
vivekp has joined #mlpack
csoni has joined #mlpack
MystikNinja has joined #mlpack
vivekp has quit [Ping timeout: 256 seconds]
vivekp has joined #mlpack
travis-ci has joined #mlpack
< travis-ci>
ShikharJ/mlpack#107 (GAN - 9098502 : Shikhar Jaiswal): The build has errored.
bhuwnesh has quit [Quit: Connection closed for inactivity]
daniel has joined #mlpack
daniel is now known as Guest68325
Guest68325 has quit [Client Quit]
dddd has joined #mlpack
dddd has quit [Client Quit]
dddd has joined #mlpack
dddd has quit [Client Quit]
daniel__ has joined #mlpack
< daniel__>
so i ve been thinking about approaches to the MVU+LRSDP functions... I think it might be a good idea to first general datasets randomly to compare both MUV+LRSDP approach and normal MUV to find out and summarise the cases where LRSDP doesn't converge.
< daniel__>
and then base on the summarised dataset , visualise it in 2d or 3d and then try to find some certain pattern that might appear in the dataset
< daniel__>
also if possible we might try to seperate the MUV+LRSDP algorithm from the api and then test it seperately
< daniel__>
in this way we can design some dataset that has certain characteristic which by pre-calculation would be garanteed to give a converging result on normal MUV / MUV + LRSDP and then just only generate dataset that has that characteristic
< daniel__>
so in this way if we find an error that would mean the bug is in the implementation of the algorithm instead of the algorithm itself...
< daniel__>
so thats what i think@MystikNinja... does that sounds practical?
csoni has quit [Read error: Connection reset by peer]
csoni has joined #mlpack
csoni has quit [Ping timeout: 260 seconds]
csoni has joined #mlpack
csoni has quit [Ping timeout: 240 seconds]
csoni has joined #mlpack
__amir__ has joined #mlpack
csoni has quit [Read error: Connection reset by peer]
< yashsharan>
Hello everyone.I was thinking of writing my GSoc Proposal related to reinforcement learning.Should I get the proposal reviewed by someone before final submission?TIA.
< zoq>
yashsharan: Hello, you can upload the application via the GSoC page, and we are trying to comment on the draft (once we have a chance).
csoni has joined #mlpack
< yashsharan>
ok thanks.I will upload it there once I finish writing it.
< zoq>
yashsharan: Sounds good.
< csoni>
zoq: Submitted a PR for the PSO p
< csoni>
zoq: Submitted a PR for the PSO optimizer and solved a few issues faced. The code is still far from complete but it helped provide insight into the API design aspects. I have a draft of the proposal (not complete yet) which I would like to share for a review, where exactly should I upload it?
< csoni>
I missed the location because of a disconnection from IRC
< yashsharan>
@zoq also should I mention the issues in the proposal which i'm working on but haven't submitted a PR yet for it?
csoni has quit []
csoni has joined #mlpack
< zoq>
csoni: on the GSoC page
< zoq>
yashsharan: yeah, that would be helpful, github name as well
< csoni>
zoq: Okay thanks!
prem_ktiw has joined #mlpack
< prem_ktiw>
Hello Everyone. I am Prem from NIT Meghalaya, India. I have done several courses on core Artificial Intelligence and Data Science such as Data Mining and Pattern Recognition in my University. I am currently in my final year and best performer of batch academically. Machine learning is an area which interests me a lot, specially the fact that things which look so fancy to outsiders are just simple algorithms dancing on the data.
< prem_ktiw>
I represented NIT Meghalaya for the first time in history of ACM ICPC Asia Regionals, my primary language for coding was C++, but that costed me to stay quite detached from Development and Open Source Developement. I have done some 5+ Projects during my college studies primarily in Java and 4 during my internship at MathWorks. if mlpack accepts me for GSoC, it will be my first open source contribution.
< prem_ktiw>
Earlier I have written C++ implementation to basic Neural Network, for one of my professor's teaching purposes. But definitely that wasn't too advanced. Menotrs please guide me about the project and further proceedings.
< prem_ktiw>
In case I am not selected via GSoC, still I would like to contribute, please let me know if that happens how to go ahead with off GSoC contribution to mlpack.
< prem_ktiw>
The first line in the section : Before you consider applying ...
< zoq>
Getting familiar with the codebase is a good first step, on that point you can build up the application, what part needs to be added/modified, what can be reused.
< zoq>
ah, thanks let me fix that
Prabhat-IIT has joined #mlpack
sourabhvarshney1 has joined #mlpack
< sourabhvarshney1>
@zoq I have submitted my draft for the Essential Deep Learning Modules. If possible, please review it and give valuable feedbacks.
daniel__ has quit [Ping timeout: 260 seconds]
sonudoo has joined #mlpack
< sonudoo>
Hello
< sonudoo>
I have not received a feedback fir my draft.
< sonudoo>
Can any mentor please review my draft and provide a feedback?
arsh has joined #mlpack
< zoq>
sonudoo: Please be patient, we will provide feedback once we have a chance.
< zoq>
sourabhvars: Okay, will take a look once I get a chance.
arsh has quit [Client Quit]
< sonudoo>
thank you. Awaiting a reply..
bhairavm has joined #mlpack
bhairavm has quit [Client Quit]
sonudoo has quit [Quit: Page closed]
qwebirc248988 has joined #mlpack
qwebirc248988 has quit [Client Quit]
sourabhvarshney1 has quit [Ping timeout: 260 seconds]
thepro has quit [Ping timeout: 256 seconds]
sujith has joined #mlpack
sujith has quit [Ping timeout: 260 seconds]
Prabhat-IIT has quit [Ping timeout: 260 seconds]
< zoq>
sonudoo: What is the title of the application/project?
yashsharan has quit [Ping timeout: 260 seconds]
< Atharva>
zoq: we don’t have an ann function to calculate the accuracy of predictions given the input amd correct classes, do we?
< MystikNinja>
daniel__: I was thinking something along those lines exactly, however I wanted to dive into the numbers and test against some existing MVU (not with LRSDP) implementation (R has it in Rdimtools, and there's an Octave implementation as well)
< MystikNinja>
Even if the same algo isn't used to solve the SDP, we should still get the same solution, no? I'll try to find out why mlpack's implementation doesn't return the same values.
< MystikNinja>
rcurtin: I'd like to know what you were thinking about this problem, where you felt the error might like and what you were doing to try to debug it
MystikNinja has quit [Ping timeout: 276 seconds]
prem_ktiw has quit [Remote host closed the connection]
Nisha_ has joined #mlpack
ImQ009 has quit [Quit: Leaving]
MystikNinja has joined #mlpack
MystikNinja has quit [Client Quit]
MystikNinja has joined #mlpack
tjtreggats has joined #mlpack
tjtreggats has quit [Client Quit]
Nisha_ has quit [Quit: Page closed]
Nisha_ has joined #mlpack
Nisha_ has quit [Quit: Page closed]
Nisha_ has joined #mlpack
< MystikNinja>
I'm trying to build an executable for mvu. Could anyone point me to where I need to make changes so that it can be compiled and built via the normal build process?
Nisha_ has quit [Client Quit]
< MystikNinja>
Do I simply have to add a line mlpack_mvu to the CMakeLists.txt file from line 486 onwards?