Learn and Practice Applied Machine Learning

Discover the self-study projects methodology you can use

Like programming, skill in applied machine learning only comes with practice. Practice on tools, algorithms and datasets. It’s hard. Where do you start and how do you use your few spare hours a week effectively?!

Over the last decade I have been using and refining a methodology of designing and completing small projects to learn, apply and implement machine learning algorithms. I used it to survive grad school, write a book of algorithms, finish client work and to do well in machine learning competitions.

I have outlined this methodology and provided a blueprint for 4 different project types called the Small Projects Methodology: Learn and Practice Applied Machine Learning.

If you want to learn and apply machine learning and like small self-study projects, then this guide is for you!

Small Projects Methodology


Applied machine learning is hard

They don’t make it easy. Machine Learning is an interesting and even thrilling subject, but the material is typically academic and dry. The focus on theory and maths can be very frustrating if you just want to try some algorithms on a dataset.

  • … where are you supposed to start?
  • … what do you do next?
  • I only have a few hours per week, how should I spend them?
  • … what are some ideas of things to try?
  • … how do you leverage what you have already learned?

Once you can use and apply algorithms practically, the theory is a lot easier to understand. You’ll have a context to understand what it’s all about.

Small Projects Methodology

Small projects is the blueprint for learning and applying machine learning, fast.

You can define your own small projects, but it can be hard to think up good structured projects with useful outcomes.

Each project is small

  • …small in scope, you have one clear objective with each project
  • …small in time, each project takes one to a few hours
  • …small in resources, you only need a PC and open source software

If you’re working full time like me, you can easily tackle one small project per week.

4 Project Types

To make this easier, I have described 4 project types in great detail.

  • Study a Machine Learning Tool: Select a tool or library that you like and learn how to use it well.
  • Study a Machine Learning Dataset: Select a dataset and understand it intimately and discover which algorithm class or type addresses it the best.
  • Study a Machine Learning Algorithm: Select an algorithm and understand it intimately and discover parameter configurations that are stable across different datasets.
  • Implement a Machine Learning Algorithm: Select an algorithm and implement or port an existing implementation to a language of your choice.

These four project types cover a wide range of applied machine learning activities from dataset investigation and visualization to applying machine learning to your own problem, getting an algorithm into operations or tackling a machine learning competition.

Use Small Projects as proof of your skills

Small projects are a powerful method that you can use at work, on competitions and to learn and practice applied machine learning.

You can use each completed project as a public demonstration of your skills:

  • …you can open source each project and list them on GitHub or BitBucket
  • …you can list each project on your LinkedIn profile or Facebook profile and link to them
  • …you can blog and tweet about the results of each project
  • …you can get comments and reviews on your results from professionals and experts

Each project is a symbol of what you have learned.

I used it to survive grad school

I was lost in graduate school. I was just a programmer thrown into the mix of research. There was so much to do and I had a lot of trouble focusing on what to do next.

About half way through my program I stopped caring about what everyone else would think and I started to design and complete small projects. I started describing algorithms at first and later moved on to designing and executing experiments. Each project was written up with an objective, results and summary of what I learned.

I refined this approach and my knowledge of the field accelerated. After months I had a directory filled with 60+ completed projects that became my dissertation.

I continued to refine the approach and used it again to complete 45+ small projects that I turned into a book on nature inspired algorithms.

Jason BrownleeAbout the Author

Who is behind this?

Hey, I’m Jason Brownlee, a father, husband, developer and author. I have written books on artificial intelligence algorithms and I have a Masters and a PhD in Artificial Intelligence.

I started out as a programmer interested in machine learning and designed and completed small projects to teach myself about the field. This lead down a path of quitting my job, studying as an AI researcher and eventually surfacing back into industry as a programmer again.

I now work in that perfect mix of developing scientific software for real users with actual problems.

I live in Melbourne, Australia and will happily talk machine learning all day long.

Order Your Copy of the Small Projects Methodology

To order your copy of the Small Projects Methodology, click the “Order My Guide” button below and then enter your payment information into the form.

We use Gumroad, a secure payment platform to collect your payment information. Your information is protected with the same encryption that your bank users. Once you pay, we’ll email you a link to download your copy of the Small Projects Methodology.


Your work has been VERY helpful for me as an aspiring Data Scientist!
— David Dalisay

Small Projects Methodology

Order My Guide
Buy with confidence! We offer a 30-day 100% Money Back Guarantee.
Reply to your purchase receipt for 100% refund for any reason!

Tabe of Contents

This exclusive machine learning guide includes…

  • Blueprint for self-studying applied machine learning
  • 4 project types described in great detail
  • 5 tactics for each small project type
  • 5-step process template for any project
  • 6 constraints to focus your efforts and get results faster

Below is an overview of the Small Projects Methodology guide including a breakdown of the tactics you can use in each of the 4 project types.

  1. Introduction
  2. Study a machine learning tool
    1. Select a tool
    2. Summarize capabilities
    3. Complete tutorials
    4. Create tutorials
    5. Create extensions
  3. Study a machine learning dataset
    1. Select a dataset
    2. Summarize the dataset
    3. Visualize the dataset
    4. Summarize the structure
    5. Run experiments
  4. Study a machine learning algorithm
    1. Research an algorithm
    2. Summarize parameters
    3. Characterize behaviors
    4. Select test datasets
    5. Run Experiments
  5. Implement a machine learning algorithm
    1. Select programming language
    2. Select algorithm
    3. Research algorithm
    4. Select test datasets
    5. Code and test
  6. Summary

Bonus Project Ideas

As an added bonus, you will get an additional bonus guide of project ideas that includes…

  • 90 different ideas for small projects
  • Project ideas listed by 3 project types (study tool, dataset or algorithm)
  • Project ideas listed by 3 skill levels (beginner, intermediate or advanced)


What if I hate the guide?
If you really hate the guide, then I don’t want your money. Just reply to your purchase receipt email within 30 days and I will issue you a refund.

What if I’m just a beginner?
Small projects is a methodology for practitioners of all skill levels: beginners, intermediates, advanced. There are suggested projects for all three skill levels and the flexibility for you to tailor your project to exactly the problems and algorithms that you want to learn about.

Where can I learn more about you?
I have tons of blog posts on MachineLearningMastery.com covering machine learning and the small projects methodology. Read a few and learn more about my teaching and writing style.

I have another question.
If you have any other questions, please contact me and I will do my best to answer them.