Hi everyone, I’m new to machine learning and looking to get started with building models. I’ve heard about both sklearn and tensorflow, but I’m not sure which one would be better for a beginner like me. I once used some basic programming languages like Python and Java. Could someone explain the main differences between sklearn and tensorflow, and recommend which one would be easier to learn and use for someone with my background? Thanks
Probably best to begin with some numbers.
To what extent are you a beginner? I think you should do neither if you’re a true beginner. You ought to become familiar with programming basics. Which language to study first is beyond me. Compared to most, Python is a little simpler to learn. Since Unity is frequently used to create games, learning C# for six months to a year might be worthwhile before moving on to learning about games.
Without getting into the difficulties of building a game (at least the kind you probably think of, or whatever you mean by framework/code, which I assume refers to web development), learning the principles is a lot of work.
Prioritizing your problems and selecting between deep learning and classical machine learning should be your first steps.
Sci-kit-learn handles classical machine learning, therefore it can be used to solve issues with sparse training data.
For deep learning, Pytorch and Tensorflow are usually used. They could be the best option if you wish to fix issues with vision or situations with a lot of data.
Yet I wouldn’t advise learning X. Learn about Deep Learning and use it with Pytorch, is my advice. Or use sklearn to apply your basic knowledge of classical machine learning.
I think Scikitlearn and TensorFlow are a bit high-level. You could use PyTorch as an alternative since you’re already canvassed with Python language. It can help you gain a bit of understanding of what all of the machine learning apis/libraries do. Once you are more familiar you could opt in for Tensorflow or sklearn
I am interested to know. Just curious, isn’t TensorFlow the same as PyTorch? I thought they are functionally the same
On the contrary, your statement is not 100% since scikitlearn is ideal for beginners. It offers deep learning on a more classical level (I’m talking decision trees and Linear/logistic regression type of deep learning models ). And you can train using small amounts of data.
Sklearn comes with many handy utilities in data preprocessing, in model selection, and algorithms (such as clustering) that Tensorflow doesn’t have . Tensorflow has a maths library, and is for neural network learning, and ofcourse as a framework it also comes with some data preprocessing tools for completeness.