Python is all about Productivity and Modularity. So it’s fairly obvious that Machine Learning will also contain certain modules that act as wrappers around complex topics, example Keras and PyTorch. Some act as modules that unlock a hidden potential, some perform complex calculations, vector operations, and Array manipulation like Numpy and others serve as a way to read datasets and load them into memory like Pandas. There are also a few Visualization libraries such as the popular MatPlotlib and Seaborn these modules allow for various charts, diagrams, representations and other such things through which humans can get the output in an easier way.
Lets see some of the commonly used Modules:
- Pandas, This library is one of the most important libraries for Machine Learning as this module handles the training data inputted by the user. What is thee use of performing Machine Learning when the data is in an incorrect format. A really good use case scenario of Pandas is when we have to read .csv files.
- Numpy, Arguably the most important library to perform complex math operations, Numpy is the beating heart of all calculations performed on a ML model. Under the hood, ML is just Matrix & Vector Manipulation.
- Scikit-learn, This library is the core of many of the ML tasks that you perform. This is where all your ML Code and Requirements reside.
- TensorFlow, This library is for advanced Deep Learning and other such tasks.
- Keras, This library is a wrapper around TensorFlow with a high level of abstraction. This library gives us the feeling that creating a Neural Network very simple.
- MatPlotlib, This library is essential to visualize your results.
These were just a few libraries but in reality we have a lot more like PyBrain, PyTorch, Tflearn, Tensorflow.JS etc.
I think that once you develop some knowledge about these commonly used modules, Learning the actual meat of ML will be a breeze.
Thank you guys, See you soon 🙂