Many people from a diverse range of backgrounds reach me on LinkedIn. Asking for tips and suggestions for transitioning their career into Data Science. I used to think, A ton of resources are available online to learn data science for free. So, Why are they asking me? After speaking to some I clearly understood,
- The abundance of options has become a curse
- Many were not sure how and where to start?
- Some were simply not confident enough
- Some were reaching out for inspiration and motivation
- Some just want to know if there is a scope for them
- Most of them were looking for a clear plan
Also going through some of the job descriptions for entry-level positions in data science. It was clear the expectations are being set very high. This is further causing a lot of stress and anxiety especially among people with non-technical backgrounds. It is also the reason why many are unable to keep up the level of motivation throughout their learning.
In this article, I am going to show you a clear plan. To help you not only learn data science but also to get a job.
Starting simple but consistent
The most common mistake made by people learning data science is starting off with a complex topic. It is natural to get attracted to an interesting topic. The problem is if you hit a roadblock too soon in your learning journey. It will be enough to drop your motivation and make you quit. The probability of running into issues is very high when you start off with a complex topic.
My number one advice to people starting with learning data science is to start small. Below are some basic skillsets that would be used on daily basis in a data science job.
- Python/R
- SQL
- Basic Stats
OK, But how much expertise is required and Where to learn? For each of these topics, I will show you reference resources from where you can learn.
Learning to code using Python/R
To become a data scientist you need to write code. From reading the data, exploring the dataset, creating a visualization, performing feature engineering, and building a model everything needs coding. There are many AutoML and No-Code ML tools coming up to automate repetitive tasks but to grow as a data scientist you need to learn to code.
If you are new to coding then web-based platforms are the best place to start learning to code. The interactive web-based environment makes your learning very easy. You would be learning more about the topic by doing it yourself. Also, there is no stress of setting up the environment and installing the tools.
One fantastic web-based platform to learn basic Python programming is Codecademy. It is a wonderful platform to learn coding in an interactive way. It is easy to continue from where you left last time and also helps to track your progress. Thus it will be good to do some learning while you are traveling or when you got some spare time. If you have decided to go with R then you could still learn here from Codecademy.
Once you complete the online tutorial for either python or R then you can try programming from your system. Install the required tools and try it out. It will be sufficient if you are comfortable with the below to start with,
- Different data types (including collection data types) and their operations
- Control statements (If Then Else, For Loop, and etc)
- Function and lambda functions
Learn SQL
The other must-have skill set for a data scientist is SQL. When you move into a corporate environment all the data would be generally stored in DBs. Unlike the learning environment, the datasets are not readily provided to the data scientists. It is the primary role of the data science team to understand and extract the required dataset. To perform that you need to have a good knowledge of SQL.
Again, the course here in codecademy provides an interactive platform to learn basic SQL. This will help you with the basics functionalities and syntax of SQL. There are online platforms like learnsql that provide practice scenarios replicating real-life challenges.
Like, python/R once you are comfortable with SQL you can install the open-source database management system MySQL. It comes with an in-built database. You can refer to my learning SQL playlist below. It would help you with the common functionalities of SQL that are used in the day-to-day job of a data scientist.
Basic Stats
The basic statistics knowledge helps in getting a better understanding of the data. As you progress and level up your skills in data science. The basic statistics knowledge will help in coming up with better solutions.
The Statistics and probability course in Khan Academy will cover all the key statistics topics required for a data scientist. If you find this course overwhelming or you are keen to get started with data science with just enough statistics then check out the below article,
It covers the basic must-know statistics concepts that are required for a data scientist. If you are more inclined towards tutorial videos then check this playlist.
Consistency
It is enough to start small while learning data science. But it is more important to ensure consistency, especially in the early days. If you learn for a week or two and take a long break then it doesn’t help with the learning. You need to dedicate a fixed amount of time every week to your learning.
Scale-up you skillset
Once you are comfortable with the basic skillsets captured above it’s time to scale up. It is time to start working on datasets and learn about data visualization, data exploration, and model building.
Learn to visualize data
Visualization is a key skill set for a data scientist. It helps in clearly communicating the insights and to bring out the patterns present in the dataset. It is easy for people to pick patterns from visual data as compared to tabular or just text data. Visualization plays important role in both data analysis and communication.
To learn more about building some amazing visualization from scratch, check out the below article.
Learning to perform data analysis
There are many data science-related courses in Coursera. It can be very confusing to choose the one best suitable for your skillset. Below are some courses you can choose based on your current level of expertise.
If you are a beginner and haven’t performed any data analysis then start with this short 2 hours course from Coursera.
If you have a basic idea about data analysis or if you have completed the above course then you can sign-up for the below course. This course has some amazing feedback and many have acknowledged getting tangible benefits due to this course.
If you are keen to learn more then you could learn about feature engineering from kaggle using the below course.
Setting up a strong foundation
Now it’s time to level up your skills. Below are some courses from Coursera that could help in building a strong foundation. It helps to understand and learn about activities performed by the data science team across the entire project pipeline. You will learn about understanding the problem, running analysis to extract inferences and insights, choosing the right model for the business problem. These courses also have projects part of the curriculum that would provide first-hand project experience.
Most of these projects are long-term, it requires at least a few weeks of effort. But these are critical and it helps in making you ready for data science interviews and challenges in the job.
Here are some details about the below courses to help you choose the one best suitable for you. Also, don’t hesitate to skim through the topics already covered.
- Option 1 – Applied Data Science with Python by University of Michigan: Below course is better suitable if you are confident about the topics covered so far in the article. It doesn’t focus on basic statistics or data analysis. It will quickly take you to the Machine Learning use cases, text mining, and network analysis
- Option 2— Data Science Specialization by Johns Hopkins University: This course is based on R. If you are learning data science using R then this will be more suitable. This is a complete data science course it covers the topics of data exploration, visualization, statistics, and model building.
- Option 3 – IBM Data Science: This is a complete data science course similar to the above Johns Hopkins. This course is taught using Python which makes it even more attractive. This course gives a better career opportunity. About 20% of those who completed this course have landed a job opportunity. This course covers the entire pipeline of a data science project. It also covers the tools, techniques, and libraries commonly used by data scientists.
Set goals and celebrate your wins
The most important part of your learning journey is to have goals and religiously tracking your goals. It is very easy to get distracted while learning data science. The best way to avoid any distractions or divergence from the original plan is to write down the goals. Most of the resources mentioned in this article have a progress tracker. Set some goals for each week and stick to them. To increase accountability share your learning progress on a blog or LinkedIn posts.
Also, sharing your achievements and progress helps in creating positive energy and motivation to pursue further. Learning data science is a long journey, you need to have your motivations high enough to sustain and complete it.
One important reason for people to quit learning data science is a lack of confidence. Having a plan and tracking the progress helps in building confidence.
Be ready for failures but continue learning
Transitioning your way into data science is not going to be a smooth one. Expect some technical issues and other possible obstacles that could shake your confidence. Just remember one thing even experienced data scientists run into issues and it is part of the career. The data science space is a fast-evolving field so it is not uncommon to run into issues. But the positive thing is there is a very good support system in place. There are a number of online communities that will help you out. Consistency plays an important role, going to bed every day with a little more knowledge will have a compounding impact on your career.
Networking
Reach out to people with a background similar to you and have found their way into data science. It helps in understanding the challenges of the job and skillsets that need to be focused on. It will give you a good picture of things you need to work on. Networking also plays a key role in the job search. A number of data science opportunities get filled directly through referrals without going to the job portals. Apart from reaching people with a similar background. Attending meet-ups and other public data science events could help in building your professional network.
Tips for getting a data science job
Many people learning data science are struggling to find a suitable job. I am going to share some tips that could help in increasing your chances of getting hired.
- Identify the companies: Playing by your strengths will give you better leverage. Identify the companies that are more aligned to your educational or professional background. There are also many generalist data science positions that you could apply for as well.
- Build a portfolio: Organizations have limited time and resources for recruiting new people. A portfolio helps in showcasing your abilities. It often increases your chances of getting shortlisted for an interview. If you are interested in building a portfolio website for free, check my article below.
How to create a stunning personal portfolio website for free
- Build a good resume: Your resume represents you and your skills. It needs to accurately represent your experience and knowledge. There are some techniques to write a better resume with impactful statements. If you are interested check this article here for tips and to learn about tools for building an impressive resume.
- Before the interview: Do your research about the role you are applying for and the company. It will be Ok to reach out to current employees to better understand the work, culture, and more.
- During Interview: Be confident and at the same time be honest about your strengths and shortcomings. Employers would prefer candidates with high integrity and lower data science knowledge over someone with very high knowledge in data science but are dishonest or not transparent.
Need some more motivation?
The below story explains the journey of the author from Neuroscience into Data Science. After 8 years of study and work in Neuroscience, the author found her way into data science.
How I went from zero coding skills to data scientist in 6 months
Below is my personal journey into data science from a sales career. I had no idea about R or Python. I didn’t have a formal education in data science or statistics. In fact, I wasn’t good at programming but now I have authored 2 books in data science and have successfully executed some turnkey data science projects.
To stay connected
- If you like this article and are interested in similar ones, follow me on Medium
- I teach and talk about various data science topics on my YouTube Channel. Subscribe to my channel here.
- Sign up to my email list here for more data science tips and to stay connected with my work
