Jobs in machine learning are very hot right now and they are consistently being voted as the best careers to enter. If you don’t have a degree then you might be wondering whether or not it is possible for you to get a job, in machine learning, without one.
This post will show you whether or not you can get an ML job without one, how to go about doing it and other things to consider.
So, how to get a machine learning job without a degree? It will be necessary for you to show that you have sufficient experience to make up for the lack of a degree. You can show that you have relevant experience by completing ML projects, doing well in ML competitions, contributing to open source projects, attending hackathons or creating your own ML blog.
Most machine learning positions will require a masters degree or a bachelors degree in a quantitative field with the ability to show relevant experience. To get a machine learning job without a degree won’t be easy especially when you will be competing with people that have degrees. For you to get your foot in the door it will be necessary for you to show a lot of relevant experience.
How to get a job in machine learning without a degree
Below, I will show you how to start if you don’t know anything about machine learning, what to do once you have learned machine learning and how to go about getting interviews.
Learn the required skills
Before you can start getting a job in machine learning it will be necessary for you to learn how to make use of machine learning.
There are a lot of skills that machine learning makes use of and many of them will be tested in coding challenges for the interviews.
To do well in most machine learning roles it will be necessary for you to have knowledge of calculus, statistics, linear algebra, SQL, programming, data structures and algorithms and of the different machine learning models.
There are two courses that I would recommend that you use to learn machine learning. The first is Andrew Ng’s course which doesn’t assume any prerequisite knowledge. The second is machine learning with Python by MIT that has prerequisites of probability, linear algebra and programming in Python.
If you need to learn calculus then I would recommend this course on Edx.
If you need to learn linear algebra then I would recommend this course on Edx.
If you need to learn probability then I would recommend this course on Edx.
It is common that you will be tested on SQL, data structures and algorithms in data science and other machine learning type interviews. So, it will be necessary for you to have a good knowledge of them and to practice them a lot on websites such as Leetcode.
Gain experience
It will be necessary for you to make up for the lack of a degree by showing that you have a lot of experience in machine learning.
Competitions
One way to get experience in machine learning is to participate in machine learning competitions.
One place that you can find a number of different machine learning competitions on is the Kaggle website. On Kaggle there are regularly ranked machine learning competitions that have prize money. If you can do well on one of these competitions you could include it as a part of your resume and it will help to show that you have a good understanding of machine learning.
With that being said, the actual process of applying the machine learning models only makes up a small part of what most machine learning jobs involve. The datasets, that you will find on these competitions, will normally be clean datasets where you will not need to do much data cleaning yourself. In a job setting you’ll likely be spending more time data cleaning than actually selecting and applying the different machine learning models.
So, it is important not to get too focused on Kaggle competitions and to make sure that you have some projects in your portfolio that show that you can clean data.
Building your own projects
One way to show that you are able to clean the data would be to build your own machine learning projects.
When making your own machine learning projects it is important to show that you have a good grasp of the whole model building workflow which includes:
- Importing the data
- Cleaning the data
- Pre-processing the data
- Doing feature transformations (creating new features using existing features)
- Feature engineering
When making your own machine learning project be sure to include the above steps. The project shouldn’t be too small either but it also doesn’t need to be too large. I would say that the ideal project to put into your portfolio would take about a month to complete. You can look at this blog post to see some examples of machine learning projects that you could make.
Once you have completed your machine learning projects be sure to post them on Github and to include a Read Me and other documentation about what you used such as the technology you used, tutorials you referenced and dependencies that you used.
Open source projects
Contributing to open source projects is another good way that you can show that you have experience in machine learning.
By contributing to open source projects you will be able to show that you are able to work as a part of a team on a large project. You will also be able to get a lot of feedback on your code from very knowledgable people on the topic which will greatly help you to improve as a programmer.
Here is a good article that talks about the benefits of using open source projects to land a job in more detail.
MyBridge shows you all of the current opensource projects, on GitHub, that you can contribute to. If you are thinking of contributing to open source projects then I would highly recommend that you take a look at it.
Create a machine learning blog
Another option that you have is to create a machine learning blog where you include links to your projects and you include blog posts where you talk about aspects of machine learning and give tutorials on machine learning. By doing so you will be able to establish yourself as an authority on the subject which will improve the chances that hiring managers will see you as a credible prospect.
Hackathons
Another option that you have is to attend hackathons. These are where people meet up for several days and code a solution to a challenging problem from scratch.
In the programming community, hackathons have actually begun to surpass job fairs in terms of showcasing yourself as a credible job prospect.
Large companies will often send their own engineers to these events to see how potential candidates perform and to see if they would be a good fit for the company. Seeing as you do not have a degree this would be a good way for you to show that you have the ability to work effectively in machine learning.
In addition to this, it will be an additional project that you can show to potential employers to showcase your ability to work on machine learning problems as part of a team.
You can find hackathons on websites such as meetup.com.
Here is a good blog post about why to use hackathons to get a job.
Here is a good video on some more ways that you can find ways to do more machine learning and data science projects:
Consider a bootcamp
Another option that you have to gain experience and to get a job would be to do a data science bootcamp. This is where you’re taught how to be a data scientist and they will often help you in finding a job after completion.
There are even some bootcamps such as the one at Springboard that come with a job guarantee at the end of it.
However, bootcamps are expensive and your prior experience and your ability to showcase that you have the necessary skills are still very important. Generally, bootcamps work well for people that have many of the necessary skills already but they are just trying to transition from one related field into machine learning.
The video below gives a good summary of what bootcamps have to offer.
Getting machine learning interviews
Once you are able to showcase that you understand how machine learning works and you are able to show that you can complete end to end projects as a part of a team then it will be time to start looking for interviews.
Go to networking events
One way to get an interview would be to start going to networking events where people meet up to talk about machine learning and aspects of data science. At these networking events you will be able to meet people in the industry that will be able to help you land a job in machine learning. Being able to get referrals from people within companies is one of the best ways to get an interview.
You can find local networking events in machine learning or data science on meetup.com.
Write an effective resume
In order to improve your chances of getting a machine learning interview, it will be necessary for you to craft your resume effectively.
You can watch the video below to see a number of tips on how to write an effective resume for machine learning positions.
Try to get an internal referral
When trying to get a machine learning position, especially without a degree, it will be very helpful to be able to get an internal referral.
You can get referrals by going to networking events, going to hackathons, contributing to open source projects, connecting with people on LinkedIn or by scheduling informational interviews.
Email recruiters directly
Another option you have is to reach out to recruiters directly. It won’t usually be as effective as getting an internal referral. However, it can work and it is generally more effective than simply applying online.
When messaging recruiters it’s important to keep the message short and it’s best to ensure that the initial emails can be replied to quickly and easily and not to straight up ask for a job in the initial email. This blog post does a good job of summarizing what to do and what not to do when emailing recruiters directly.
Apply online
Another option that you have is to apply online.
Online applications will usually get hundreds of applicants so your chances of getting an interview from one are slim especially if you don’t have a degree. With that being said, it is still worth trying.
Prepare for the interview
Data science interviews will normally include questions on mathematics, statistics, SQL, data structures and algorithms.
There are many ways that you can practice for what will be asked on data science interviews.
One option is to practice SQL, data structures and algorithms using Leetcode and to practice the maths and statistics problems using Brilliant.
You could also read the book Heard in Data Science Interviews which goes through 650 of the most common data science interview questions and their answers.
Also consider freelancing
Once you have gained some experience by creating your own machine learning projects, attending hackathons, doing machine learning competitions and contributing to open source projects you could also consider getting work by freelancing.
There are sites such as UpWork.com that will allow you to find people looking for freelance. However, the work on there will generally not be the best work.
It will be necessary for you to initially start by working on the less desirable tasks possibly at a lower pay rate. But, after you have done a few gigs you will be able to find work much more easily and you will be able to increase your rates.
You can watch the video below to see a good overview on how to go about being a freelancer in AI.
Consider what your current skills are
There are a number of different types of machine learning jobs that you can get and each of them will require different types of skills.
If you have strong software engineering skills then it would likely be better for you to consider becoming a machine learning engineer. The role of machine learning engineers is to take the machine learning models and to make them usable, at scale, in a production environment.
If you have strong math and stats skills then it would likely be better for you to consider becoming a data scientist.
If you have good analytics skills then becoming a data analyst might be your best option. Seeing as you are trying to get a job without a degree then you will likely find that data analyst positions will be more accessible to you since companies tend to be more willing to hire people without degrees for this position. Once you are working as a data analyst then you could work on transitioning into more of a data science type role.
Be willing to get told no
When applying to the different machine learning positions you will be competing with people that have degrees. Many hiring managers will prefer to hire people that have degrees. When they have an option to either interview someone with a degree or someone without one they will usually choose the person that has one. This is why it will be necessary for you to be ok with being told no.
With that being said, if you are able to show that you have a lot of experience then, if you keep applying to the various openings, then you’ll have a good chance of getting offered interviews and hopefully positions.
