IDEAS
This talk is presented by Zach Miller, Senior Data Scientist at Metis
Learn more: https://www.ideassn.org/
Linkedin us: https://www.linkedin.com/company/data-science-association/
Like us: https://www.facebook.com/datasciencea…
Follow us: https://www.facebook.com/ideassn/
Instagram us: ideassn
Source
added to my favorites!
He makes a very good point towards the end for those who are average or less with Math. Imagination and creativity with data is a big area and I can guarantee that you will find Math and non Math people who are good in this area. In some ways this is the most important area because coming up with models that are better than Joe's is usually because you thought of a different way of looking at the data or maybe you felt that some combination or feature engineering would produce better results.
sense of humor much appreciated!
you are the boss of the boss
12:06 my OCD is going off the charts. He wrote overflow bellow stack, even if there was enough room to write it above.
Why use two screens and then the presenter has to be out of the frame once in a while. I think that's dumb!
The information is quite useful!
I like everything he said except he is saying it takes few weeks or few months to get certain things done. No it takes way more effort than that unless you are damn smart
Informative one, and the key skills are statistics and programming for machine learning
thanks for uploading this, appreciate it
Can I learn all of this by myself ?
Hii.. This is aditi Jain and I am doing PhD from finance and I want to make career as a data scientist.. So please tell me what should I learn first
http://Www.threelly.com
I started the same way. I am a physicist and nanotechnologist, did PhD in electrical engg. Now working as a professor in machine learning and AI
Um is not a verbal period.
I'm still 17 years old
the Algorithm Selection image at 12:49 was worth the price of admission.
asians …comeon..learn to speak english atleast. i couldnt understand their questions properly .imagine trying to work with someone who has speech impediment.
Thank you for this man.
Great talk Zach,
Good site he mentioned:
https://machinelearningmastery.com
Error in rcorr(cor.government) : could not find function "rcorr"
please what causes this?
hope he didnt get payed for it
You can't. Machine Learning is not microwave popcorn or minute rice.
this genius is a part of the mega hadron collider project !!! man ! they have the best minds on the planet ! damn !
Please cameraman, do a better job next time and follow the instructor. Thanks
@14:11 note for self.
fucking void
I have bachelors degree in PHYSICS, MATHEMATICS AND INFORMATION TECHNOLOGY..i know programming languages like c,c++,java and sql…. should i continue to have masters in physics or should i switch to ML…. interested in both
한국에서는 SAS나 무슨 H…어쩌구 저쩌구를 마니 쓰는데 미국에서는 python 하고 R, JAVA 학교에서 마니 가르침 학교수업 거의다 Rㅎㅎ 빨랑 복습해야딩
Slides – http://zwmiller.com/blogs/ZWM_6MonthsToMachineLearning.pdf
I can't find Stats and probability course videos on mit open course. I found other 2 course but not this. Please help me find it.
while i really liked this talk, i have a problem with the following:
from "6 months on your own with absolutely zero private time" to "12 weeks that you can leisurely take because its all set together for you already"
doesn't sound right to me
Can I have cheat sheet at 13:10 ? it too small on video
Is this where you learn all the math: calculus, linea algebra, stats, and then here: scikit learn and tensorflow for everything and forget the math
What domain is scope, either rpa or mean stack to start a career
I only wonder how u can keep on these boring things in 6 months without any motivation… might be the most important question for me… 🙁
Fuck.. It's very difficult when your English is bad but you must to understand this roadmap. I need translator 🙂
Thank you very much !!
Timeline is bullshit. I've got a BS in Astrophysics and no way it will only take someone 2-3 months to learn all the maths to a point where he/she can use it easily especially Linear Algebra and Calculus. Probability and Stats is one thing, I'm better at Calc than Linear but for someone to learn all those in 2-3 months is just stretching and overselling it. In addition, if you want to learn the basic of Python starting from scratch, after 1 month you probably will only get the basic and will be unable to know it well enough to be able to use it easily. Just on math and learning python good enough will take much longer than this 3 month projection.
I use …lot of data of all engineering branches….to build ..60 mw power plant ..single hand .. without any help….plus… process..sugar plant ..up to 10000 tcd…we create data..to..build machine s….
Pure PHD in any scientific..deciplane…without..any..bottom up.. industrial.. experience…that is.. to…say
Build.. scientific.. complex.. plants or.. machines..not.. by..single hand…but..with team works…is ..just..piece..of.. toilets… paper..!!!!!
best 23 minutes of my life . thanks a lot
a lot of this is no longer required unless you're making your own model which there's a ton out there
I started a week ago and I think I will require 1 year to complete (with 6 hours a week).
PYTHON or R? I started with Matlab and OCTAVE
can we get the ppt