Data science is another name for the science of data. It is a field of analysis that has developed worldwide at an astonishing pace in recent years. Data science owes this development to the enormous growth of the Internet and the social-networking services operating within it, such as LinkedIn and Twitter.
There's a lot of retail transactions online, there's information services, and there's huge agencies running unprecedented-scale advertising campaigns.
Every second, billions of clicks are made on servers around the world, and every client's choice, thanks to the great computing power, is carefully recorded and analyzed.
For large-scale Internet companies, every customer's behavior, every customer's correctness is carefully analyzed and recorded in that customer's account.
If we've been searching the Internet lately for a vacuum cleaner, the vacuum cleaners will be following us, and wherever we go, there's always going to be a discreet commercial for the vacuum cleaner, and if we hate celebrities and we love football, then the algorithms of the news media will be pushing the latest football news and isolating us from the scandals of unpopular celebrities.
What counts is how long we keep a customer on the site, how often they come back to us, and by adapting to their needs, the websites make them dependent on each other, giving them a competitive edge.
W-MOSZCZYNSKI-2019-1-01What does this have to do with the bakery?
The example of data science activity on the Internet is not very relevant in the context of bakeries and confectioneries, and I've used this example because this activity is very visible to everyone, and even though you can't see it, data science works in all areas of the economy.
Statistical process analysis has developed particularly where mass operations are carried out, in cellular networks, gas and electricity distribution, in banks, in retail networks and in logistics.
Data science in my bakery?
Let us leave the noise and clamour of the wider world and return to the quiet of our bakery. Here, too, changes have taken place in production and distribution technology. Nevertheless, the entire sales process, the production cycle and the handling of raw materials have remained unchanged for years. Is it worthwhile to introduce innovations from another world in such an environment? The idea seems absurd!
What is data science really?
This field of knowledge was not invented yesterday. It has existed for more than 100 years, concealed under different names. Seventy years ago it was statistics; in the 1960s it shone as econometrics and mathematical optimization, only to appear in the early 1980s as the then-fashionable Six Sigma. Each time, the statistical revolution foundered on the rocks of laborious calculations and cumbersome systematic work. Analytical work still had to be conducted, but patience and widespread knowledge were lacking. Every time, something was missing that would allow the analytical undertaking to take on a life of its own.
The digital bakery?
What makes the bakery and candy from the 1920s different from the bakery from 2019 is that the latter, in addition to flavoring, produces huge amounts of digital data. The tax bill generates dozens of information on a one-time sale, such as price, transaction time, assortment indexes, the name of the salesperson.
If there's a few stores, in total, there's thousands of transactions every day, so let's look further, we have hundreds of data from computers that control gas stoves, we have data from recipe systems, we have inventory data, we have data from car computers, and finally we have data from inventory returns.
This last piece of information tells us how much and which products didn't sell, so something was wrong, we made the wrong decision, and now we have to pay for it.
The vast amount of data from the digital bakery collected here and now is stored on computer disks, even without our knowledge. It just has to be, it's a requirement of the Treasury Department, it's a service register of devices or a memory of logs of sales and production programs that we don't even know about.
Every digital image of the day here and now combines with the previous days to create endless series of time data.
The digital world outside is influencing the bakery: currency rates, gas prices, and the weather, the approaching holidays and the change in time to summer have a direct impact on sales volume, technical cost of production and product returns.
This data is a treasure trove of information about demand, consumer behavior, production and distribution efficiency, and the interdependence of phenomena is obvious to the naked eye, you know something is coming from something, but we don't know exactly what, we don't have time to analyze, we don't have time for statistics, we suffer because we don't know.
The situation is reminiscent of when a small yacht crashed on a tropical, uninhabited island, and two months later most of the wrecks died of starvation and exhaustion, and the island was rich in vegetation and fresh water, and the sea was full of fish, and none of the wrecks could use these resources.
A candy store with statistics?
And what's different about bakeries in the 1920s from bakeries in 2019 is the technology of data processing, and the baker used to keep trading books, pencil operations, and run the calculations on the margins, and the problem was that every time our baker had to run all the calculations over again.
Now, in the digital revolution today, it's not just that we've got the formulas in Excel that we're going to replace the data and get the result.
It's like the baker is surprised that every time he gets into a delivery car, he doesn't have to reassemble the engine of that car.
There are companies that would do anything to know how much raw material they have to order to produce, there are banks that would do a lot to know how their customers are behaving in response to a tariff hike, and finally there are candy companies that would like to know how much they have to produce so that they don't have to waste those products at night.
Information about the future has always been important, and today information about the future is becoming available.
Who is the Data Researcher?
Data Scientist Data researcher is the highest-paid job in the industry today. There are many types of data scientists, and some of them are full-time, and they're most often bank fraud detectors, process optimizers, and process controllers.
Most Data Scientists, however, are nomads, so-called digital nomads, involved in short, well-paid technology projects.
In their case, the purpose of such projects is to create self-learning forecasting models or intelligent, autonomous optimisation systems.
They leave behind completely autonomous, unserved learning systems that efficiently and free of charge provide key information about the future, and the lifetime of the models they create is unlimited and depends solely on the characteristics of the process they're living in.
Until the technology of production, distribution or sales organization changes, the model algorithms will work forever.
In the next few episodes of this series, I'll show you some practical applications of data science in bakeries and confectioneries, and I'll also show you how to build simple predictive models in Excel yourself.
Wojciech Moszczyński — graduate of the Department of Econometrics and Statistics of Nicolaus Copernicus University in Toruń; specialist in econometrics, finance, data science, and management accounting. He specializes in the optimization of production and logistics processes. He conducts research in the area of the development and application of artificial intelligence. For years he has been engaged in the popularization of machine learning and data science in business environments.
Dodaj komentarz