This device is not compatible.

You will learn to:

Understand Bayesian probability theory.

Perform Bayesian inference from scratch.

Use Matplotlib to visualize how the models work.

Learn how Bayesian methods can be applied to ranking.

Skills

Machine Learning

Data Science

Data Visualization

Prerequisites

Proficiency in probability and statistics

Good understanding of Python

Familiarity with NumPy

Technologies

SciPy

NumPy

Matplotlib

Project Description

New and rarely seen items often present difficulties in data science. Even though data is available in larger quantities today than ever, dealing with sparse data is as important as ever. As the number of users has grown, so has the number of things they can interact with—consider the sheer number of videos on YouTube, for example. Not only that but as the amount of data grows, we often try to push it to its limit by analyzing ever more fine-grained segmentation.

Comparing items with different amounts of data can be tricky. In this project, we will use the Bayesian probability theory to solve this and related problems mathematically principled. We will apply Bayesian inference to multiple datasets, using data visualization throughout to show how the models work and understand their outputs.

Project Tasks

1

Getting Started

Task 0: Introduction

Task 1: Import the Libraries

Task 2: Understand Bayes’ Rule

Task 3: Understand the Binomial Distribution

2

Estimate Conversion Rates

Task 4: Create the Conversion Rates Dataset

Task 5: Compute Maximum Likelihood Estimates

Task 6: Compute a Baseline Using Additive Smoothing

Task 7: Determine the Distributions

Task 8: Use Empirical Bayes

Task 9: Compute Bayesian Estimates

Task 10: Plot and Compare

3

Estimate Purchases over Time

Task 11: Create the Dataset for Purchases Over Time

Task 12: Calculate Maximum Likelihood Estimates

Task 13: Find the Distributions

Task 14: Compute Bayesian Estimates

Task 15: Plot and Compare

4

Application: Ranking Products

Task 16: Investigate an Application to Ranking

Task 17: Discuss an Alternative Method

Congratulations!