Experience, ah that golden part of a resume. Good to have especially when its relevant and often not having it can be why you didn’t get the job. And for people trying to break into a new field or fresh out of school, it very naturally leads to the question ‘if no one will hire me, where am I supposed to get experience?’ If you’re an aspiring Data Scientist, the solution can be to build a project.
Data Science education opportunities are wide in scope, including formal basic and advanced degrees, certificate programs, and a variety of online learning formats. Reputable education providers are the best way to learn the basic and advanced skills data scientists need to succeed. But what it doesn’t provide is real hands-on experience. There are a few internships available (IBM is advertising one, exactly!), which certainly will help you get real world experience. They are truly limited – IBM is advertising one – yep exactly one! Plus, for many aspiring data scientists an unpaid or even a low paying internship is not an option, they need to keep their current jobs while they work towards a career in data science.
This is where having your own project can prove to potential employers what you can create. Projects can focus on just about anything, Data Optimal suggests these five types of data science projects will boost your portfolio, and help you land a data science job:
1. Data Cleaning is probably the most in demand action a data scientist can perform. If we know nothing else about data, we know that lots of it is just a mess. Some of this is caused by the overwhelming volume of data and, that first time efforts don’t always get the data as clean as needed to be useful. Here is a great example of real-life data cleaning project by Tich Mangono. A quick Google search can supply lots of sources for data that could use some cleaning.
2. Exploratory Data Analysis allows for a greater understanding of data and does it in a visual way. Generating questions using this process can lead to new, insightful discoveries. For the stakeholder a visual display of the data and how it will help solve their problem is the ultimate objective for a business. Being able to produce actionable data is the true value of data analysis. This interesting project on the cost of missed doctor’s appointments by William Koehrsen is a good EDA example.
3. Interactive Data Visualizations take data analysis a step further using tools like dashboards. Corporate oriented end-users and the data science team can work together to draw insights that shape business decisions. Plus, for end-users that may very possibly be overwhelmed by the plethora of detail a dashboard is an easier way to work with the data. Microsoft provides a useful guide for creating dashboards.
4. Machine Learning sounds like the basis for a futuristic science fiction movie, in this case it is relying on patterns or inference vs. specific requests. Start with the basics such as linear and logistic regression which are easier to explain to upper level management. Focus on a project that has business impact and you’re sure to impress potential employers. Denis Batalov’s project using AWS, does a great job of exploring the impact of customer churn – a very important business issue.
5. Communication has become more and more important as the major skill needed in every business, in every field. Regardless of how insightful or valuable your model may be, you must be able to explain it to the stakeholders for buy-in. Communication is what makes a great data scientist versus a good one. Take a look at these data science presentations on Data Science Central using SlideShare for some inspiration. Keep in mind that presentations are relevant, geared toward the specific audience and demonstrate results that have a business impact. And as the old saying goes ‘practice, Practice, PRACTICE.’
Document all your projects on GitHub using a GitHub Pages portfolio and link them to your LinkedIn profile so they can be found easily by potential employers. GitHub provides a plethora of tools and services in an environment with other data scientists. Share your work, get input and showcase your projects.
Once you’ve got a project or a few to show potential employers you should find that the experience part of your CV is now an asset. You should be able to knock that next job interview out of the park!