You do not have any formal education in data science but you are interested in venturing into this exciting field that everyone is talking about. So how easy (or difficult) is it to find the perfect data science learning path for you?
What if you are too lazy to comb through all blogs and platforms to find that perfect path? Do you just want to follow an idiot but a foolproof path to get into data science? Here, we answer just that and present to you the best learning paths we found online.
These are our criteria in selecting learning paths;
- Covers end-to-end machine learning process
- Covers both programming and math fundamental of data science
- Provides appropriate certification or credentials (to get a job!)
First up – Datacamp Career Track
One of the most popular data science online learning platform. Datacamp has an amazing collection of 321 courses produced by an army of 272 instructors who are constantly releasing new content. Learners can choose to take these courses separately or as part of a skill or career track. As of right now, they are boosting an incredible 40 different types of skill tracks and 11 career tracks to cater to your every need. Courses consist of short video lectures, finger exercises, and interactive programming environment that provide a wholesome experience for learning. You will not need any additional tools or download any IDE to work on Datacamp, they had it all prepared for you!
Quoting from Datacamp, career tracks cover everything you need to kickstart your career, while skill tracks are shorter and let you specialize in a specific area. For example, learning to import and clean data are part of a skill track whereas learning to become a Data Scientist with R is one of the career track. In mathematical terms, skill track is a sub-set of career track!
If you opt for Datacamp, your biggest issue will be selecting the career tracks you wanted to invest in. Although you can change track anytime after you have started on one, its better if you choose and focus on one. Some pointers for beginners, 1) Choose either Python or R, do not jump between programming language, 2) Focus on career track, these are structured curriculum that will cover most skill tracks needed, 3) Apart from video lectures and interactive assignment, make use of projects and start building up your portfolio.
If you need more reasons to start learning Data Science using DataCamp, we have a dedicated post just for that. Read more here.
Pros and ConsStrong Community Large Collection of Projects Customizable Learning Path Subscription-based
Next up – Dataquest Career Path
Ah the million-dollar question. Datacamp or Dataquest?
Our answer is the one that suits your needs.
Very similar to Datacamp, Dataquest also has a skill and career track implementation. Although courses weren’t as extensive as Datacamp, it is definitely sufficient to get you in the data science field. Since it shares a lot of similarity with Datacamp, we will list their main differences;
- NO video lectures. Theories and concepts are all explained in text format and only suitable for someone who prefers reading than watching lectures.
- No independent projects. Projects are incorporated into individual courses instead of a separate entity like Datacamp.
- Certifications are per-course basis. You get certificate for each complete course but not upon completion of career track (Major downside)
Pros and ConsCustomizable Learning Path Suitable for learners who prefer reading Lacking in advanced topics Subscription-based
Advanced Learning path – MicroMasters in Statistics and Data Science
If you follow us long enough, you will realize our data scientists love the MicroMasters Program by MITx, and with great reasons.
The Statistics and Data Science MicroMasters Program is a collection of 5 courses that span across 1 year and 3 months. Completing the 4 required courses + 1 virtually-proctored capstone exam will lead to a MicroMasters certificate that counts as credits towards a Master’s program in more than 10 universities, including MIT.
The program consists of a series of graduate-level courses namely;
- Probability – The Science of Uncertainty and Data
- Data Analysis in Social Science – Assessing your Knowledge
- Fundamentals of Statistics
- Machine Learning with Python: from Linear Models to Deep Learning
- Capstone Exam in Statistics and Data Science
1 – 4 are advanced-level courses that will require various prerequisites and serious time commitment. They are meant for motivated learners who wish to dive deep into the respective area of data science and build a career in data science. Do not underestimate these courses as they follow the exact curricula of MIT on-campus courses, differentiating them from other MOOCs. All are instructor-led courses, meaning that homework, assignments, and programming exercises followed strict deadlines with exam dates fixed. Definitely plan your schedule before committing to these courses.
5 is a capstone exam that tests your understanding of the previous 4 courses. You can only take the capstone exam once you completed 1-4 as verified learners and obtained the certifications. The capstone consists of 4 timed virtually-proctored exams that you must take and get an average score of 50% to pass the MicroMasters Program.
Comparing to other MOOCs like Datacamp and Dataquest, the MicroMasters Program aim to provides learners with enough depth rather than breadth. It train your ability to think analytically and ensure a fundamentally sound process for your data science projects.
This MicroMasters is a great data science learning path to close the gaps between self-taught data scientist and those that went for formal university/postgraduate education.
Pros and ConsUnbeatable, high quality contents Similar pace and level of rigor as on-campus courses Recognized Certificates Not for beginners!
Special Mention – Udacity NanoDegree
Technically, Udacity’s Nanodegree does not fit our criteria as you need to purchase multiple nanodegrees to cover the whole machine learning processes. However, it is still worth mentioning on this list.
Udacity packaged their courses into multiple nanodegree programs that aim to teach a specific skill set in each program. But most importantly, each nanodegree grant you access to their career services to help you with your job hunt.
- Personal career coaching
- Interviews preparations
- Resume services
- Github review
- Linkedin profile reviews
making it one of the best platform that focuses on getting you employed. You should really make full use of the services and to network and advance your career.
Since there is no one nanodegree that cover the whole data science process, here are some recommendation;
|1||Programming for Data Science with Python||If no prior programming experience|
|2||Intro to Machine Learning with Tensorflow||Choose between Tensorflow or Pytorch|
|2||Intro to Machine Learning with Pytorch||Choose between Tensorflow or Pytorch|
|4||Computer Vision||Advanced topics|
|4||Natural Language Processing||Advanced topics|
Having a learning path is only one part of becoming a Data Scientist. Click here for our detailed guide to become a Data Scientist in 2020!