Great talk and as always it was very well presented (both speak and slides)! As I'm not a maths guy at all, I would have expected some pieces of code in the slides.
what a way to wake up in the morning :-) a rather mathy talk, would have loved to see a bit more code but then the formula is probably easier to understand than a bit of php code that implements it.
Heavy for a day after slot but excellent talk. Andrei explained the math background rather well, the use case was interesting, attractive slides and all well presented. Very slick.
Realy nice introductiong into machine learning using a real world example. Really enjoyed this talk, might be interested into using some similar techniques in upcoming development.
How's that for a wake-up call :) Great to see yet another "different" talk at a PHP conference. I liked the topic and it was very well presented. I liked the fact that an actual use case was used to clarify the topic.
If I had to put in one bit of critiscism: I don't mind the formula's, I would have liked it even better if they too would have been "filled" with the actual parameters so they illustrated their own use and meaning better.
Best talk of the conference! Interesting stuff, illustrated with a real world example. There might have been some more code examples in the presentation. However, Andrei made an example available on Github, which you can find here: https://gist.github.com/4642209. A big thank you for that!
The presentation itself was well structured with great slides. Well done!
Enjoyed it although I got lost in the math examples a few times. But the nice examples, slide design and your style of presenting it helped me to get on track again every time. Some live examples would have made it perfect for me.
Very interesting, a lot of math that I still have to get my head around. One of the talks that really stood out from the rest, and the proof that there's so much more cool stuff to be learned ;)
Some hands-on code examples would have been great, though.
missed a little bit of details. how did you choose alpha (step size) in optimization problem. and how did you choose number of cycles for gradient descent ? did you randomly initialized starting points (initial thetas).
did you wrote optimization algorithm in php or did you used sklearn for this.
did you tried SVM and libsvm extension for php ?
but overall it inspiring talk to start with machinelearning
Comments
Comments are closed.
Great talk! I really enjoyed it thoroughly. Great slides too!
– @michieldewit
nice talk, some simple explanation for some more complex math examples
You managed to explain complex formulas on screen in understandable terms. It was an interesting problem and solution you presented, very enjoyable!
Great talk and as always it was very well presented (both speak and slides)! As I'm not a maths guy at all, I would have expected some pieces of code in the slides.
what a way to wake up in the morning :-) a rather mathy talk, would have loved to see a bit more code but then the formula is probably easier to understand than a bit of php code that implements it.
Very interesting. Makes me want to do more math.
Good speaker too.
Heavy after the social, bit very cool use case.
Great start of the day, will certainly look into this for something I recently started on!
This was the best talk of the conference for me! thanks a lot
Heavy for a day after slot but excellent talk. Andrei explained the math background rather well, the use case was interesting, attractive slides and all well presented. Very slick.
Realy nice introductiong into machine learning using a real world example. Really enjoyed this talk, might be interested into using some similar techniques in upcoming development.
A real good introduction to machine learning. Too me the best talk of the conference.
How's that for a wake-up call :) Great to see yet another "different" talk at a PHP conference. I liked the topic and it was very well presented. I liked the fact that an actual use case was used to clarify the topic.
If I had to put in one bit of critiscism: I don't mind the formula's, I would have liked it even better if they too would have been "filled" with the actual parameters so they illustrated their own use and meaning better.
Best talk of the conference! Interesting stuff, illustrated with a real world example. There might have been some more code examples in the presentation. However, Andrei made an example available on Github, which you can find here: https://gist.github.com/4642209. A big thank you for that!
The presentation itself was well structured with great slides. Well done!
Enjoyed it although I got lost in the math examples a few times. But the nice examples, slide design and your style of presenting it helped me to get on track again every time. Some live examples would have made it perfect for me.
Very interesting, a lot of math that I still have to get my head around. One of the talks that really stood out from the rest, and the proof that there's so much more cool stuff to be learned ;)
Some hands-on code examples would have been great, though.
Definitely one of my favourite talks of the confernce. Great hands-on examples, good slides and brought really wel. Great job!
missed a little bit of details. how did you choose alpha (step size) in optimization problem. and how did you choose number of cycles for gradient descent ? did you randomly initialized starting points (initial thetas).
did you wrote optimization algorithm in php or did you used sklearn for this.
did you tried SVM and libsvm extension for php ?
but overall it inspiring talk to start with machinelearning