Data analysis is a very large section since there is so much information to digest. However, there are 4 steps to make it easier.
Finding Patterns: First off, finding patterns help identify important data items that can be analyzed. Repeated items stick out and can be further explore. For example, if a certain search item was constantly in a user’s search history, it might give you a clue as to what the user likes.
Classification: Classification is a step in which you look at certain chunks of the data, instead of the entire data set. This narrows it down so it can be analyzed with ease, but also focuses on the most important parts of the data. For example, say you want to find out a good vacation spot, and you are part of the middle class. You might want to take a survey of a couple of families to see where they last went for vacation, but with all the data, you might only use the middle class families’ answers. This is because those vacations would be the most cost effective for your status.
Association: Association is a technique that is most used in marketing. Basically, you put items that are closely related near each other in a store, and you figure out what items are closely related through patterns. This technique is good for learning about buying trends. For example, paper cups and plates are often bought together, so you put them near each other in a store.
Prediction: Finally, prediction is a technique that is used to find out the relationship between two variables. This goes along with association. For example, you might find that whenever someone buys a sleeveless dress, they might buy a cardigan along with it. Therefore you put the cardigans near the dresses. This way even if some customers do not need a cardigan, they might see it and get attracted, leading to them buying one anyway. You are basically predicting the relationship between buying cardigans and dresses.
References: https://neebo.ai/data-mining/
Comments