Machine learning expands the boundaries of what's possible by letting software accomplish tasks that can't be accomplished algorithmically. In this hands-on workshop, learn the fundamentals of machine learning from a software developer's perspective and learn how to use it to build sophisticated predictive models. You'll learn about regression, classification, recommender systems, and more, and you'll do it with Scikit-learn, the world's most popular machine-learning library. You'll also see examples of machine learning in action and build and train models of your own.
Prereqs:
Our day will be divided into four sessions, each consisting of a lecture and a set of demos followed by a hands-on lab. The labs are an important part of the learning process, because YOU get a chance to practice what’s presented in the lectures. Please plan to bring your laptop with you for the labs. You’ll also need to prepare your laptop with an environment for working the labs. Fortunately, it’s easy. There are a couple of ways to go about it:
- OPTION 1: Install the 64-bit version of Anaconda (https://www.anaconda.com/). It provides a complete Python machine-learning environment and includes important packages such as Scikit-Learn.
- OPTION 2: Install a 64-bit version of Python 3.8.5 or higher (https://www.python.org/downloads/) if it isn’t installed already. Then execute the following commands to install the Python packages that the labs require:
- pip install notebook
- pip install scikit-learn
- pip install pandas
- pip install matplotlib
- pip install seaborn
If you are unable to perform either of these installations, an alternative is to use Colabs, Google's free implementation of Jupyter notebooks in the cloud. Colabs offer a friction-free way to run Jupyter notebooks with common Python packages such as Scikit-Learn already installed. They also include free GPU support and require nothing to be installed on your computer. To get started, go to https://colab.research.google.com/notebooks/intro.ipynb and sign in with your Google account. If you don't have a Google account, you can create one for free.