Diving Deeper into Machine Learning

We’ve focused on neural networks, using labeled data that we can use to learn the trends in our data. This is an example of supervised learning.

Broadly speaking there are 3 main approaches to machine learning

• Supervised learning

  This uses labeled pairs (input and output) to train the model to learn how to predict the outputs from the inputs.

• Unsupervised learning

  No labeled data is provided. Instead the machine learning algorithm seeks to find the structure on its own. The goal is to learn patterns and features to be able to produce new data.

• Reinforcement learning

  As with unsupervised learning, no labeled data is used, but the model is “rewarded” when it does something right, and the model tries to maximize rewards (think: self-driving cars).

Libraries

There are a number of popular libraries that implement machine learning algorithms. Their features and performance vary quite a bit. An comparison of their features is provided by Wikipedia: Comparison of deep learning software.

Some additional comparisons are provided here: https://ritza.co/articles/scikit-learn-vs-tensorflow-vs-pytorch-vs-keras/

• TensorFlow

  This is an open source machine learning library released by Google. It has support for CPUs, GPUs, and TPUs, and provides all the features you need to build deep learning workflows: TensorFlow feactures.

• PyTorch

  This is a machine learning library build off of the Torch library, originally developed by Facebook.

• scikit-learn

  This is a python library developed for machine learning. It has a lot of sample datasets that provide a nice means to learn how different methods work. It is designed to work with NumPy and SciPy.

  General recommendations on the web seem to be to use Scikit-learn to get started with machine learning and to explore ideas, but to switch to one of the other packages for computationally-intensive work.

  Scikit-learn provides some nice sample datasets:

  https://scikit-learn.org/stable/datasets/toy_dataset.html

  as well as generators for datasets:

  https://scikit-learn.org/stable/datasets/sample_generators.html

There are also tools that provide higher-level interfaces to these

• Keras

  Keras is built on top of TensorFlow and provides a nice python interface that hides a lot of the implementation details in TensorFlow.

Keras / TensorFlow

We’ll focus on Keras and TensorFlow.

There are a large number of examples provided by Keras:

https://keras.io/examples/

You should be able to install keras via pip or conda.