Zen and the Art of Data Science by Helge Reikeras

The popularity of Data Science in the business world has exploded in recent years as companies are realising the value that data can yield to their products, services, and business decisions.

In this talk, we'll discuss a typical data science workflow, from extracting the raw data through to serving real-time machine learning predictions using a REST API.

We'll feature a range of Python tools that make up the Data Science pipeline, including Pandas, Scikit-learn, Gensim, Luigi, and Flask, and show how these all can work together.

A real-world implementation done at Offerzen will be presented as an example.

Transcript

2. • • • • •
7. cookiecutter https://github.com/drivendata/cookiecutter-data-science
8. class HelloWorldTask(luigi.Task): task_namespace = 'examples' def requires(self): return [FooTask(), BarTask()] def run(self): with self.output().open('w') as f: f.write('Hello World') def output(self): return luigi.LocalTarget( '/data/hello_world.txt')
9. from flask import Flask app = Flask(__name__) @app.route('/') def hello_world(): return 'Hello, World!'