In my pastime, I like to watch videos, read articles, or even just check the news about all things new in the tech world. Whether it’s innovations or every just privacy update like what’s been going in with iPhone, it’s always interesting to stay on top of what’s going on. When I think about keeping up with data science it’s interesting to stay up to date on improvements no matter what field of study that it comes from. Sometimes it may be Machine Learning, other times it may be AI (artificial intelligence), or my personal favorite, AR (augmented reality). I’m especially interested in data visualization, especially through AR.
So far nothing too ground-breaking. But today I want to talk more about how all-encompassing data science can be. The inspiration for this article is my number one fan, which is my girlfriend who reads all my articles even if she doesn’t have any interest in "data science", which she categorizes as Big Data and machine learning. Although both are a part of data science, today I want to talk about how important data science can be to all fields of study, especially those in Computer Science/Engineering and related fields, from Dev Ops to Analysts to Developers, and everyone in between. Even the business side can take a big help with data science updates and innovations, as it is much more all-inclusive in terms of data. So, without further delay, let’s look at why I believe data science is important for everyone to keep up with.
The first type of analytical Data Science: Prediction
When you think of machine learning, you are likely thinking about Predictive analysis. This is where an algorithm is created to recognize patterns so decisions can be made based on those observations. But if you’re not all for machine learning, why does Predictive analysis in Data Science still apply? Predictive analysis is relevant whether done by a machine or done by hand. And that is with more occupations than only technical. For example, a doctor would look at family histories, which are recognizable patterns, to determine if you’re at risk for certain diseases. Or a small baker may see their sales statistics to determine early Autumn is the best time to bring out their pumpkin spice recipes.
While you can use your own eyes to make these predictions, machine learning seeks to make this faster. Instead of surveying a thousand people in a population, for example, business analysts could find a sample of that population, then allow machine learning to predict the answers of the remaining members. This could save large amounts of time, as what would have taken weeks can be completed in a few minutes. Time saves money, making breakthroughs to become faster at predictions essential to running a relevant business.
But even in your everyday life, the predictive analysis appears more than you may think. When you type a message on your phone, autocorrect may fix your spelling mistake, or your phone may even guess the next word you want to write. Even extensions such as Grammarly can be used to predict what you want to write. So, you see, now you can younger generations that back in your day you didn’t have any fancy word correctors or predicters. You had to fix those mistakes yourself. These updates are important, like with Grammarly, because everyone in College could use the extension to clear up their grammar mistakes and make their wording more professional without having to find a tutor or proofread again and again.
As a developer specifically, maybe that prediction could help you. In terms of only the text, extensions such as Intellisense could help to finish your thoughts when coding. Such as selecting the right method or finding the table you were halfway through typing the name of. This was brought up because even if you’re not a fan of learning Tensorflow or any other machine learning language, that doesn’t mean you can’t appreciate the world around you.
The second type of analytical Data Science: Description
Descriptive analysis in data science refers to the comprehensive data, and any means used to visualize that data. This could refer to monthly statements, graphs, presentations, or even more techy approaches such as dashboards. Of the four types, this is the most common. This is because not only can developers use it, but it is also key to business users. Without the representation of data, it would be difficult to find meaning in it. For example, if a business is selling a certain number of books per month, they could use graphs to visualize whether their techniques are improving, staying the same, or slowing down. For at-home developers, maybe you are attempting to find a solution to sort photographs. You are considering a few different types of coding methods and test around three of them. You could use a graph or even a dashboard to visualize what percentage of images were sorted correctly with each method. This would help you to both improve your methods, but also select which approach you should likely take.
Data Visualization is critical to businesses. R is a platform specifically for that visualization in data science. R is also able to represent big data. This makes visualizing data much faster and easier than manually crunching the numbers.
The third type of analytical Data Science: Diagnosing
Another type of analytical data science to discuss is diagnostic analysis. As it may sound, this is finding the root of a problem. It may involve both drilling down and isolating all confounding information. The goal of the diagnostic analysis is to find the issue faster so a solution can be made. In our previous dashboard example, this may represent finding data that is causing issues, then drilling into the details of that issue.
For a developer, the dashboard example is still relevant. However, we can use a more all-encompassing example. In your chosen IDE, I’m sure there are times where you’ve typed something wrong or called a method that doesn’t exist. Maybe even used the wrong case in a variable declaration. You may notice that familiar red squiggly line. The IDE recognizes there is an error, but they have gotten smarter about recognizing where that error is. So, for a lot of newer versions, the days of searching two hours for a missing semi-colon are over. Now, your IDE can find what line is throwing the error. Some even provide hints of what you may have meant to write instead.
The fourth type of analytical Data Science: Prescription
The final type of data science we’ll look at is prescriptive analysis. The reason I am explaining this one last is that it needs at least some, if not all, of the others to create a solution. That means that the prescription, which you can probably guess, is the solution to some given issue. Being able to visualize, and therefore describe the issue, would be important to know how to fix it. You would also need to diagnose the issue to form a solution. You may even need to find solutions to other predicted events. So, as you can guess, they’re all linked closely.
The prescriptive analysis uses a combination of machine learning and statistical models to recommend strategies based on test outcomes. The more advanced techniques are used, the more specific the recommendations can be.
To explain this by hand, imagine you want to have a coding session (or just a hangout) with a group of friends. To meet, you need to exchange schedules to see when you would all be available at the same time. By linking each other’s calendars, you can prescribe which day to meet, which would ensure no one has overlapping plans. You could do that either by hand or over the computer with your calendars. But how would that example overlap with bigger portions of data? In college, notice how sometimes your final exams are kind of at weird times. But 9 times out of 10, you’re thrilled that none of them seem to overlap. This is because colleges can use your schedule, represent the data of other students in the same class, and prescribe a day and time (even location) that wouldn’t conflict. Now, that system isn’t always perfect. But that’s because perhaps some of their predictions were a little bit off. But that’s just all part of the learning process.
Another real-world example, one that you probably use often, is how your GPS can predict, then prescribe to you, the fastest route to get to your location. It can measure the distance and compare road speeds to determine which route would be the fastest. Innovations are making it even easier to predict where traffic jams occur and how long it will take to get around them. Some systems have even started allowing you to report accidents or speed traps. So really, in your everyday life data science is still quickly helping you to make decisions and solve some issues.
Put the Science back in Data
Maybe it goes without saying, but without science, your data would be meaningless. From the number of sales to the fastest time on the Mario Kart lap, unless you make observations and comparisons, the data would be worthless. But as developers, we rely on data science more heavily than just worrying about our jobs having successful business ventures. Even when trying for a new job, the data on your resume is used and compared to others, and decisions are made with that data. Whether you’re interested in big data or more comfortable at a distance, computers are making decisions faster and easier, but also making helpful predictions about the market, consumer trends, or even just what the weather will be like based on other patterns observed.
Conclusion
No matter what field you’re looking into, and especially for fields in technology, data science has become an ever-increasing factor that everyone should know at least a little bit about. No, you don’t have to dedicate your free time to studying new trends as I do. But now and then, it doesn’t hurt to just see what’s happening in the data science world. As we discussed, there’s much more to data science than just machine learning. There’s a little bit in data science for everyone, even non-tech fields. Even in your everyday life, you find instances of machine learning aimed to make your life easier.
For developers, it becomes even more important to keep up with updates. One example we talked about briefly is data visualization, which is essential to everyone in a business, but even aspects like prediction and prescription can be incredibly useful, especially for developers. Tools like Intellisense can make coding much easier, or at least time-saving. And having your IDE of choice prescribe exactly what line of code your error is on, and even suggestions on how to correct those errors. Even if you just enjoy the benefits of machine learning, keeping up with the trends and updates can at the least paint a clearer picture of what you can expect in your coding future. At least, in my opinion, keeping up with data science can be both interesting and insightful. I hope you enjoyed my opinions about how important machine learning can be for all fields. Until next time, cheers!
Read all my articles for free with my weekly newsletter, thanks!
Want to read all articles on Medium? Become a Medium member today!
Check out some of my recent articles:
A Simpler Introduction to OAuth
References: