The main goal of statistical data analysis, that is, data science, is to understand data. Once you understand it, you can also control it and manage it to meet your present and future goals. Including business goals.
Statistical data analysis allows us to find relationships, associations, trends and patterns in data. Things that you cannot see with your naked eyes. But the powerful tools of statistics can. From the classic and basic statistical methods like simple linear regression analysis to the new and advanced methods like support vector machines.
If just 20 years ago a feature film like "Toy Story" caused sensation because of its realistic computer-generated imagery, now that same looks basic compared to today's standards. The same huge advances in computing power has enabled the implementation of very powerful statistical techniques that weren’t just possible before.
What is understandable is also controllable
These techniques work through a few or dozen or hundreds of millions of instances in any kind of data you have, each single instance having just a few or one hundred thousands of attributes of any kind of data. Sizes and dimensions that human analysis can’t cope with, can’t see through. But that well tested and proved statistical tools, based on solid mathematics, can. Now the implementation of these techniques are possible, thanks to the enormous advances in computing power in the last two decades, especially in the last 10 years.
A wide range of powerful statistical techniques and models enables the understanding of data and its control in very diverse areas. Their applications are numerous. Here we will list just a few. These techniques can help to find out:
- which products are commonly bought together so that target discount and marketing displays and materials can be customized, bringing large increase in sales.
- how much an increase in advertising of a product in one media will impact its sales, how much advertising in more than one media will have a positive inference effect on sales.
- which chemical properties of a product (wine, perfume, compounds) bring a better quality to focus only on the best ones
- which machine configurations lead to production failures to anticipate machine stops and increase production greatly
- which medical expenses incurred by insurers will have a major impact in the overall medical expenses and which kind of customers are more likely to have the bigger expenses and where
- which kinds of loan applicants are more likely to default
- which kinds of patients or diseases will require more facilities and more human resources allocated in a medical institution like a hospital
And many more.