+ 7

Getting into machine learning

I wanted to participate in a deep learning project at my university, but there was no chance because it required experience. However, I'm still very interested in it. I guess that Google's TensorFlow library for Python is the most famous tool for this matter. The professor leading the project also recommended Keras, an API which makes TensorFlow more user-friendly. Does anyone here have experience with machine learning already? Is it difficult, is it exciting? Do you even have some projects you could show or tell about? :)

9th Feb 2018, 7:23 PM
Chris
Chris - avatar
4 Antworten
+ 11
It is an awesome field to bind your future with! You get to try out so many applications of mathematical and statistical theorems in practice and amazingly they do their job as well as you wouldn't have imagined :) I took part in a six-month data science bootcamp during which I couldn't wait to learn more and more. I am still unfullfilled seeing how much there is to learn with me having so little time between my work, private and family life and getting some sleep :) Anyway, drop by to see my github account at: https://github.com/kuba-siekierzynski and take a peek at some of my projects, especially the CNN one. I used keras and tensorflow to build an image classifier, which was a fairly easy to do, but harder to fine-tune and increase prediction score of. Still, it was fun and educative, I put it in a form of a Jupyter notebook, so that anyone can take it and run it on his own :) Have fun! 🐍
9th Feb 2018, 9:14 PM
Kuba Siekierzyński
Kuba Siekierzyński - avatar
+ 6
Data Science & Machine Learning is great. You can start Machine Learning here: https://www.kaggle.com Have fun!
9th Feb 2018, 7:55 PM
Elias Papachristos
Elias Papachristos - avatar
+ 3
I'm just curious. When you went and told them you wanted to participate, despite the "required experience," what did they tell you as a response to your question? Personally, I haven't dealt with machine learning, but I've researched the subject and it certainly sounds fun/exciting. It's a good indicator of the future, so it's worth knowing / learning. However, I don't base anything that I do on difficulty because whether it's hard or easy, if it's what I want to do then I'm going to tackle it head-on, know what I mean? When I label things in a difficulty type of way, it changes the perception of what's going on.
9th Feb 2018, 7:31 PM
Fata1 Err0r
Fata1 Err0r - avatar
+ 1
(bookmarked)
10th Feb 2018, 4:35 AM
何岱庭
何岱庭 - avatar