Data Science is a big buzzword right now. But what is it? And where do the likes of Machine Learning and AI come in?
This week, we thought we’d grab some time with our resident data enthusiast, undercover jazz pianist and all-star educator, Dan Adler to find out why Data Science and Machine Learning matter.
Hello! Who are you and what were you up to before Coder Academy?
Before Coder Academy, I was working in the Advisory Analytics group at PwC in the U.S. ‘Advisory’ is just another name for consulting…which also happens to be another term for helping companies and people find solutions to tough problems. ‘Analytics’ involves using math, models and computers to help analyse data. In short: we helped companies solve tough problems by analysing their data.
A lot of the problems I worked on were in healthcare, specifically population health. We helped hospitals and medical insurance providers understand their populations’ behaviour outside of their environments l by combining things they know about their patient population with data they didn’t have, like consumer trends, or population statistics. We’d then help them understand the causes of chronic conditions within their patient populations, and — more importantly — understand how they could invest in their communities to lower the burden of diseases in the future.
And we did this with math…a LOT of math.
What brought you to teaching at Coder Academy?
I’ve always been really passionate about finding creative ways to garner people’s interest in math, science and engineering. Since university, I’ve been blogging (shameless plug) about how math relates to things we experience on a daily basis, and I’ve been teaching people (of all ages) over the past couple of years in engineering and computer science.
I think that a lot of the changes we’re seeing around the world — whether it’s movements towards automation, climate change, or even cyber security, it’s becoming more and more important for people to get involved in technology. A recent report from the Brookings Institute mentioned that 70% of jobs in certain sectors are at risk to be automated, which is a HUGE number. And while politicians continue pointing fingers at other potential sources of these economic shifts, we need to face the actual problem by making sure we open up pathways for everyone to up-skill or retrain into a career in tech.
That’s what brought me to Coder Academy. Our staff truly cares about promoting equality and opening up opportunities inside tech to people from all backgrounds. Across all of our courses we have diverse students each undergoing their own unique journeys — from mothers returning to work, to school leavers who are looking to creatively jump start their careers. Additionally, we try to lower the barriers that society sets so that it’s easier to reach one’s own potential.
How would you describe the education model here at CA?
Every time we approach a course or a classroom, we set up the framework for a curriculum and then actually engage in a conversation with our students to really make sure that we’re giving them the best learning experience possible. This is something you only learn by being in the classroom. No matter how much planning you do, you always need to be agile and willing to respond to change. I can say that 99% of what we do is try and empathise with our current students so that we can support, challenge and help them create incredible things in short spaces of time.
There’s a ton of stuff people need to learn to be successful as a developer and inside the workplace— not all of which are the hard technical skills. You need to be able to give presentations on the spot, run meetings and write well. Our Bootcamp experiences are all about practical projects and team collaboration — it’s intense, but the results have really been incredible.
The tech industry is all about Data Science right now. But what is it, really?
It’s funny because the formal field of “Data Science” has become the IT buzzword recently, but personally I think data science has been around for a long time.
Data Science is the science of creating processes to analyse data. I would say that anyone who analyses data in excel is doing data science. Sure, maybe it’s not some huge complex analysis, but you are analysing data to make a decision…which sounds like data science to me. The reason why we’ve started to focus much more on data science recently is because the amount of data we can hold and process has grown exponentially. Now that we can process more data, we’re starting to have a feedback loop where we can do a lot more interesting things.
There’s a really awesome article that explains more on this from the NY Times using Google Translate as a case study — it’s a total must read if you’re interested in this space.
Is Data Science important?
Data Science is becoming very important. More and more people are starting to use data and machines to automate processes instead of having humans do tasks themselves — it’s becoming part of our daily existence.
We can see this in simple examples, like people ordering Big Macs using a machine in their local McDonalds, to more complex examples where businesses like THE ICONIC are starting to use machine learning to study customer behaviours and optimise targeted advertising. Machines like this are learning more and more each day, and they’re increasingly involved in much of our lives.
So yes, we need to pay attention to data science. We need to make sure we have people who know how to control the things we build so that these machines are making the right decisions for us. It’s really exciting, but really scary at the same time.
There’s also a lot of talk surrounding automation and machine learning across industries, is there a difference between Machine Learning and Data Science?
I would consider machine learning (ML) a subset of data science. Data science involves everything in the process of using and analysing data: from storage, to visualisation and analysis. ML is really more focused on just the analysis portion. ML is the process of “teaching” a machine to find patterns in data. These patterns might be things we know and can explicitly tell a machine, like how the French word “après” translates to the English word “after”, or things that we do not explicitly tell a machine, but want a machine to figure out itself.
Recent reports from firms like McKinsey have estimated that 43% of jobs in Australia will be automated by 2030. What do you think the reality of automation looks like for future and current workforces?
That’s a tricky question. As you can tell, I am really stoked about the potential that machines have to utilise data science and change how we work. At the same time, it scares me that as a workforce we are not necessarily prepared with the skills to build and regulate this technology. Plus, there are a lot of people at risk of losing their current jobs because of automation so we need to face this head on and open up pathways for people to get educated in skills relevant for the jobs of the future.
Why should people get involved in Machine Learning and Data Science?
Really for the reasons above — ML and data science are changing the workplace. We can either be proactive to make sure we re-skill the current workforce, or panic ten years from now when these changes have already taken place.
Where do ethics come in?
Diversity — and not just demographic diversity, but a really diverse set of thinkers — allows us to be more conscious of the implications of our technology. When we build these systems that utilise ML and data science, these systems are only as good as the data we put into them. If that data is biased in some way, these biases will translate to the decisions are machines make.
You can read a really good example of bias in a natural language processing tool commonly used called Word2vec. In short, when a computer tries to find patterns in text data, it cannot actually understand text, so it tries to translate text, and subject matter within text, to rows of numbers, and Word2vec provides this translation mechanism.The end result of this translation is something called a “word embedding”, which codes the relationship between words.
Now, here’s an example of how these tools can reflect some really bad biases. If you ask the machine this question: “man : computer programmer :: woman : x”, the machine will respond, “homemaker”, which is incredibly sexist and terrible. To dig deeper, these translators are built on reading Google News articles, so these translators are only biased because we as a society are biased.
Who would make a good data scientist?
Usually people who are logical thinkers, great communicators, have a business mindset, can understand patterns, know a little bit of math, and are able to understand relationships or connections between different things. I’m not saying one person has all of these skills, but there are different areas within the data science community that someone who has 2–3 of these skills could serve.
What advice would you give to someone considering learning more about these areas of tech?
There are a TON of different ways to start learning — resources like Coursera, and Kaggle, and — of course — our very own Coder Academy Machine Learning and Data Science short courses. Read and pick at everything out there, explore different viewpoints and discussions — there’s so much you can do with machine learning and data science, the world is really your oyster.
Any last pearls of wisdom you’d like to share?
Coder Academy is the most loved Australian Coding Bootcamp provider according to Course Report. Our accredited coding Bootcamps will set you up for a new career in tech via our Fast Track Bootcamp or our Flex Track Bootcamp . Our immersive courses help students acquire in-demand skills through hands on, project-based training by industry educators and experts.
Now enrolling | domestic & international students in Sydney, Melbourne & Brisbane! Study now, pay later with FEE-HELP!