Select Page

5 Ways To Gain Real-World Data Science Experience

Gaining data science experience without having a data science job seems daunting. One of the biggest questions I get from people trying to break in to data science is “how do I gain data science experience if I don’t have a data science job?”. To answer this question, I’ve put together the following top 5 ways to gain useful, real-world, data science experience:
  • Build Small Projects
  • Volunteer as a Data Scientist
  • Join a Meetup
  • Create Tutorials
  • Contribute to Open Source Projects
I’ll go through each one of these in detail, and give you specific actions to take. You certainly don’t need to do everything on this list before you can start applying for data science jobs. My advice would be to focus on one or two things in this list, don’t try to do everything at once.

Build Small Projects

My first, and probably favorite method, for gaining real-world data science experience is to build small projects. I say “small” because it’s not a complete end-to-end project. If you try and accomplish too many things at once, you won’t accomplish anything. You’re better off picking one specific skill, making a small project, and showcasing it on your GitHub Pages portfolio. github pages Here’s some examples of quick projects you could work on.

1. Collect and Clean a Messy Dataset

Collecting and cleaning messy dataset is a great way to showcase the following real-world skills:
  • How to scope a project and plan for collecting data
  • Working with different data formats (CSV, JSON, XML)
  • Different data collection methods (SQL, APIs, web-scraping)
  • Thinking through different data-cleaning tasks
The best way to create a data cleaning project is to find a messy dataset, pick a tool, and focus in on a skill. data cleaning example Here’s some sources of messy datasets:

2. Explore a Dataset

Exploring a dataset, also know as exploratory data analysis (EDA) is a great way to showcase your ability to do the following:
  • Form relevant questions
  • Investigate those questions
  • Utilize different types of plots
  • Present your findings
histogram data visualization Here’s some sources for performing an exploratory analysis:

3. Make Predictions

I would recommend creating a data cleaning or exploratory project before a machine learning project. Machine learning is by far more popular, but you’re likely to spend more time cleaning and exploring data in a real-world data science job. That being said, it’s still important to understand the basics of machine learning. I’d recommend putting together either a logistic regression, or linear regression model. logistic regression These are great models because they’re much easier to interpret than something like a random forest. It’s vital to understand the basics before trying to overcomplicate things. Here’s a couple of datasets to get you started: Make sure to document all of your work, and write it up neatly. A great tool to use is Jupyter Notebooks. This will allow you to write and execute your code, while also being able to document your work using Markdown. jupyter notebook Once you’ve written up your project, be sure to document your work on your github profile as well as your github portfolio page. Here’s some additional resources for building data science projects.

Volunteer as a Data Scientist

Volunteering your time as a data scientist is another great way to gain some experience. Another benefit is that you’re contributing your time for the better good. One option for getting started with Data Science volunteering is with DataKind. DataKind You can either volunteer to join a project, or you can submit your own project. Here’s some of the fields in which DataKind is tackling critical humanitarian issues:
  • Education
  • Poverty
  • Health
  • Human rights
  • Environment
  • Cities
DataKind has a ton of awesome projects to get involved with. Here’s a sample of just a few: DataKind Data Science Project Another great option is DrivenData. This company hosts online competitions focused on social impact. Its similar to Kaggle, but rather than focusing on business impacts, the efforts are for social good. DrivenData Some of the areas that DrivenData holds competitions in are:
  • International development
  • Health
  • Education
  • Research and conservation
  • Pulic service
DrivenData has some really interesting competitions. Here’s a sample of some present and past competitions: data science competition Volunteering on data science projects for social good is also great for building your teamwork skills. A big part of data science is being able to communicate and work with others. Being on data science team as a volunteer is a great way to develop those skills and grow.

Join a Meetup

Meetups are another great way to gain some real-world experience in data science. These are events that are held in person in your local city. They can range from educational presentations to team-based coding projects. meetup Meetups are great because they also provide great networking opportunities. You can get access to real-world data scientists from companies of all different sizes and backgrounds. To find your local meetups just do a quick internet search on “your city” + “data science meetups”. I did a quick search on New York, and here’s what I found: data science meetups These meetups range from educational to informative. For example, the “Data Science for All: New York” meetup’s goal is “…to prepare attendees for doing data science as well as expand the set of skills of practicing data scientists.” Some of their past meetups include the following:
  • An introduction to Generalized Linear Models
  • Data Science Project Night
  • Linear Regression in Detail
At the Data Science Project Night you can get advice on and work on projects. If you don’t have a project, you can work with others on their projects. data science project night Events like this are a great way to learn from others. Both about technical skills like coding and machine learning, but also how to be a better communicator and team member. The “Scala + Python + Spark + Data Science” meetup provides hands on coding practice. Some of their recent meetups include:
  • Basic Python Games – Sharpen Your Python Skills By Building Simple Console Games
  • Introduction to Natural Language Processing in Python
  • Intro to Applied Statistics for Data Science
Meetups are really an awesome way to get some hands on experience, and learn from others.

Create Tutorials

Teaching a skill and showcasing it in a blog post is another awesome way to gain experience. You’ll showcase to others that you can perform the skills that you’ve advertised on your resume. It also showcases your communication skills. In the workplace you need to be able to be a good listener as well as a good teacher. Here’s a few ideas for different types of tutorial type projects:
  • Teach a useful skill – how to clean a messy data set
  • Knowledge skill – how to interpret the coefficients of a logistic regression model
  • Advanced skill – write an algorithm from scratch (how does backpropagation work?)
I would start with doing a tutorial on a useful skill or knowledge skill. A good useful skill to teach is something like, “how to detect missing values with R”. detecting missing values with R You could find a messy dataset and then walk through your process of detecting messy values. Be sure to show your code along the way, and explain what functions you’re using, and what exactly they’re doing. My blog post on Data Cleaning with R is a good example of this type of project. A knowledge skill is another great type of project, especially if you’re learning something. Maybe something like “interpreting coefficients of a logistic regression model”. logistic regression coefficients My favorite way to learn something fast is to try and teach it. Have you ever thought you understood something, and then you try to explain it to someone, and realize you can’t? Try taking a topic and writing a blog post on it. Really explain the details, but try not to use a lot of technical jargon. Being able to understand a topic and explain it in simple detail is a truly valuable skill that employers want to see.

Contribute to open source projects

The final way to gain data science experience is by contributing to open source projects. This is a great way to get familiar with popular data science libraries. Here’s a sample of some open source projects in data science: contributing to open source Contributing to open source will get you familiar with version control and git. If you don’t have any experience with git, not a big deal, here’s some great resources to get you started: learn git Contributing to open source projects can seem intimidating, but keep in mind you don’t necessarily need to contribute code. Open source projects are always looking for contributions to their documentation. For example, freeCodeCamp is always looking for contributions to their programming guides. If you want to get started contributing, check out freeCodeCamp’s contribution guidelines. contributing guidelines for open source Open source project will usually have some sort of contribution guidelines, so it’s good to read through those. Here’s the contributing guidelines for the open source project I mentioned previously: If you’re still feeling overwhelmed, no worries, check out these beginner guides for getting started with open source projects: Contributing to open source will not only increase your coding and software skills, but it will teach you how to work with others. You’ll be constantly communicating with people running the project, as well as others contributing in the open source community. This is a great way to learn from others, while contributing to your efforts to really amazing open source projects.

Conclusion

Getting experience as a data scientist without having a data science job can be intimidating. The best way to do it is to dive right in and start applying data science skills. In this article we went over the following 5 ways to gain useful, real-world, data science experience:
  1. Build Small Projects
  2. Volunteer as a Data Scientist
  3. Join a Meetup
  4. Create Tutorials
  5. Contribute to Open Source Projects
Start by just focusing on one, and be sure to document your work on you LinkedIn profile and your github portfolio page. These projects will help you learn valuable data science skills, as well as other important skills such as communication and teamwork. If you want some more ideas for gaining data science experience, check out my guide on data science projects.

5 Ways To Gain Real-World Data Science Experience

Gaining data science experience without having a data science job seems daunting. One of the biggest questions I get from people trying to break in to data science is “how do I gain data science experience if I don’t have a data science job?”. To answer this question, I’ve put together the following top 5 ways to gain useful, real-world, data science experience:
  • Build Small Projects
  • Volunteer as a Data Scientist
  • Join a Meetup
  • Create Tutorials
  • Contribute to Open Source Projects
I’ll go through each one of these in detail, and give you specific actions to take. You certainly don’t need to do everything on this list before you can start applying for data science jobs. My advice would be to focus on one or two things in this list, don’t try to do everything at once.

Build Small Projects

My first, and probably favorite method, for gaining real-world data science experience is to build small projects. I say “small” because it’s not a complete end-to-end project. If you try and accomplish too many things at once, you won’t accomplish anything. You’re better off picking one specific skill, making a small project, and showcasing it on your GitHub Pages portfolio. github pages Here’s some examples of quick projects you could work on.

1. Collect and Clean a Messy Dataset

Collecting and cleaning messy dataset is a great way to showcase the following real-world skills:
  • How to scope a project and plan for collecting data
  • Working with different data formats (CSV, JSON, XML)
  • Different data collection methods (SQL, APIs, web-scraping)
  • Thinking through different data-cleaning tasks
The best way to create a data cleaning project is to find a messy dataset, pick a tool, and focus in on a skill. data cleaning example Here’s some sources of messy datasets:

2. Explore a Dataset

Exploring a dataset, also know as exploratory data analysis (EDA) is a great way to showcase your ability to do the following:
  • Form relevant questions
  • Investigate those questions
  • Utilize different types of plots
  • Present your findings
histogram data visualization Here’s some sources for performing an exploratory analysis:

3. Make Predictions

I would recommend creating a data cleaning or exploratory project before a machine learning project. Machine learning is by far more popular, but you’re likely to spend more time cleaning and exploring data in a real-world data science job. That being said, it’s still important to understand the basics of machine learning. I’d recommend putting together either a logistic regression, or linear regression model. logistic regression These are great models because they’re much easier to interpret than something like a random forest. It’s vital to understand the basics before trying to overcomplicate things. Here’s a couple of datasets to get you started: Make sure to document all of your work, and write it up neatly. A great tool to use is Jupyter Notebooks. This will allow you to write and execute your code, while also being able to document your work using Markdown. jupyter notebook Once you’ve written up your project, be sure to document your work on your github profile as well as your github portfolio page. Here’s some additional resources for building data science projects.

Volunteer as a Data Scientist

Volunteering your time as a data scientist is another great way to gain some experience. Another benefit is that you’re contributing your time for the better good. One option for getting started with Data Science volunteering is with DataKind. DataKind You can either volunteer to join a project, or you can submit your own project. Here’s some of the fields in which DataKind is tackling critical humanitarian issues:
  • Education
  • Poverty
  • Health
  • Human rights
  • Environment
  • Cities
DataKind has a ton of awesome projects to get involved with. Here’s a sample of just a few: DataKind Data Science Project Another great option is DrivenData. This company hosts online competitions focused on social impact. Its similar to Kaggle, but rather than focusing on business impacts, the efforts are for social good. DrivenData Some of the areas that DrivenData holds competitions in are:
  • International development
  • Health
  • Education
  • Research and conservation
  • Pulic service
DrivenData has some really interesting competitions. Here’s a sample of some present and past competitions: data science competition Volunteering on data science projects for social good is also great for building your teamwork skills. A big part of data science is being able to communicate and work with others. Being on data science team as a volunteer is a great way to develop those skills and grow.

Join a Meetup

Meetups are another great way to gain some real-world experience in data science. These are events that are held in person in your local city. They can range from educational presentations to team-based coding projects. meetup Meetups are great because they also provide great networking opportunities. You can get access to real-world data scientists from companies of all different sizes and backgrounds. To find your local meetups just do a quick internet search on “your city” + “data science meetups”. I did a quick search on New York, and here’s what I found: data science meetups These meetups range from educational to informative. For example, the “Data Science for All: New York” meetup’s goal is “…to prepare attendees for doing data science as well as expand the set of skills of practicing data scientists.” Some of their past meetups include the following:
  • An introduction to Generalized Linear Models
  • Data Science Project Night
  • Linear Regression in Detail
At the Data Science Project Night you can get advice on and work on projects. If you don’t have a project, you can work with others on their projects. data science project night Events like this are a great way to learn from others. Both about technical skills like coding and machine learning, but also how to be a better communicator and team member. The “Scala + Python + Spark + Data Science” meetup provides hands on coding practice. Some of their recent meetups include:
  • Basic Python Games – Sharpen Your Python Skills By Building Simple Console Games
  • Introduction to Natural Language Processing in Python
  • Intro to Applied Statistics for Data Science
Meetups are really an awesome way to get some hands on experience, and learn from others.

Create Tutorials

Teaching a skill and showcasing it in a blog post is another awesome way to gain experience. You’ll showcase to others that you can perform the skills that you’ve advertised on your resume. It also showcases your communication skills. In the workplace you need to be able to be a good listener as well as a good teacher. Here’s a few ideas for different types of tutorial type projects:
  • Teach a useful skill – how to clean a messy data set
  • Knowledge skill – how to interpret the coefficients of a logistic regression model
  • Advanced skill – write an algorithm from scratch (how does backpropagation work?)
I would start with doing a tutorial on a useful skill or knowledge skill. A good useful skill to teach is something like, “how to detect missing values with R”. detecting missing values with R You could find a messy dataset and then walk through your process of detecting messy values. Be sure to show your code along the way, and explain what functions you’re using, and what exactly they’re doing. My blog post on Data Cleaning with R is a good example of this type of project. A knowledge skill is another great type of project, especially if you’re learning something. Maybe something like “interpreting coefficients of a logistic regression model”. logistic regression coefficients My favorite way to learn something fast is to try and teach it. Have you ever thought you understood something, and then you try to explain it to someone, and realize you can’t? Try taking a topic and writing a blog post on it. Really explain the details, but try not to use a lot of technical jargon. Being able to understand a topic and explain it in simple detail is a truly valuable skill that employers want to see.

Contribute to open source projects

The final way to gain data science experience is by contributing to open source projects. This is a great way to get familiar with popular data science libraries. Here’s a sample of some open source projects in data science: contributing to open source Contributing to open source will get you familiar with version control and git. If you don’t have any experience with git, not a big deal, here’s some great resources to get you started: learn git Contributing to open source projects can seem intimidating, but keep in mind you don’t necessarily need to contribute code. Open source projects are always looking for contributions to their documentation. For example, freeCodeCamp is always looking for contributions to their programming guides. If you want to get started contributing, check out freeCodeCamp’s contribution guidelines. contributing guidelines for open source Open source project will usually have some sort of contribution guidelines, so it’s good to read through those. Here’s the contributing guidelines for the open source project I mentioned previously: If you’re still feeling overwhelmed, no worries, check out these beginner guides for getting started with open source projects: Contributing to open source will not only increase your coding and software skills, but it will teach you how to work with others. You’ll be constantly communicating with people running the project, as well as others contributing in the open source community. This is a great way to learn from others, while contributing to your efforts to really amazing open source projects.

Conclusion

Getting experience as a data scientist without having a data science job can be intimidating. The best way to do it is to dive right in and start applying data science skills. In this article we went over the following 5 ways to gain useful, real-world, data science experience:
  1. Build Small Projects
  2. Volunteer as a Data Scientist
  3. Join a Meetup
  4. Create Tutorials
  5. Contribute to Open Source Projects
Start by just focusing on one, and be sure to document your work on you LinkedIn profile and your github portfolio page. These projects will help you learn valuable data science skills, as well as other important skills such as communication and teamwork. If you want some more ideas for gaining data science experience, check out my guide on data science projects.
Free Bonus: Top 12 Data Science Books for 2018

Free Bonus: Top 12 Data Science Books for 2018

You have Successfully Subscribed!

Top 12 Data Science Books for 2018

Top 12 Data Science Books for 2018

Sign up with your email address to get instant access for free!

You have Successfully Subscribed!

BONUS: Data Science Experience Quick Start Guide

BONUS: Data Science Experience Quick Start Guide

Sign up for FREE to gain instant access to the Data Science Experience Quick Start Guide!

You have Successfully Subscribed!