Data sets are files created by collecting and storing a particular subject as a collection of numbers or values. For example, the files formed by collecting and storing the test scores of each student in a certain class, the license plates of the vehicles , their colors, and their speeds in a specific tabulation frame are a data set. The data set corresponds to the contents of a single database table or a single statistical data matrix; where each column of the table represents a specific variable and each row corresponds to a particular member of that data set.
As an example, in the graphic below, there is a section from the “December 2020 Traffic Density Data” provided by IBB Open Data Portal.
Artificial intelligence is algorithms written to give the machine the ability to think and act by imitating the working logic of human brain. These algorithms need to see and describe events and objects in the outside world, just as the human brain does when learning. However, people do this over many years, and the information and images that have been accumulated over these years in their mind can only be transformed into useful and usable instruments at the end of this process. In artificial intelligence technology, it is possible to obtain meaningful outputs from these data with some mathematical calculations by instantly giving the previously collected and processed data to the system. By looking at the traffic density data from IBB given as the example above, the estimation of how the density might be in the future can be made with a certain margin of deviation. Or suppose we are trying to build an autonomous vehicle, when we give the artificial intelligence of the vehicle almost flawlessly photograph and action information, the system can make an inference about how it should react at that moment by looking at the video images it takes instantaneously.
There is a strong and indispensable relationship between artificial intelligence algorithms and data sets. Artificial intelligence cannot make meaningful and accurate inferences in instant situations without large amounts of data prepared properly. For this, there are companies that specialized in preparing data sets and some government institutions.