Descriptive analysis is used to understand the past and predictive analysis is used to predict the future. Both of these concepts are important in machine learning because a clear understanding of the problem and its implications is the best way to make the right decisions. In this article, I’ll give you an introduction to descriptive and predictive analytics in machine learning.
Descriptive and Predictive Analysis in Machine Learning
Descriptive and Predictive Analysis are types of statistical analysis techniques structured as a sequence of steps that you need to take to gain comprehensive domain knowledge to solve complex business problems. These techniques give you a clear understanding of the business problem so that you can make the right decisions. Let’s take a look at descriptive and predictive analytics in machine learning one by one.
Before using a machine learning algorithm, it is very important to acquire abstract knowledge of the problem. The goal of descriptive analysis is to find an accurate understanding of the problem by asking questions from historical data. Let’s understand the descriptive analysis process using an example. Suppose your task is to optimize the supply chain of a department store, for this task we have purchase and sales data. After analyzing the data, we make assumptions that sales increase during the day just before the weekend. This means that our machine learning model is based on periodicity. So, descriptive analysis helps us understand the deep patterns from the data to uncover all those special features that were overlooked at the initial stage.
In short, the purpose of descriptive analysis is to enable us to understand whether the machine learning model will perform poorly or whether it is the best model in a particular problem.
Predictive analytics is an important concept in machine learning. What happens is that once we have formed a machine learning model based on descriptive analysis, the next goal is to infer its future steps by giving some initial conditions. Predictive analytics is used to discover and define certain rules that underlie a process for pushing a particular condition on time. For example, the object detector of a self-driven car can be extremely precise at detecting an obstacle in time, but another model must take action that minimizes the risk of damage and maximizes the likelihood of safe movement.
Predictive analytics, therefore, means observing a problem in time and taking the most appropriate action as a prescription to avoid any type of risk.
The purpose of descriptive analysis is to enable us to understand whether the machine learning model will perform poorly or whether it is the best model in a particular problem. Whereas predictive analysis means observing a problem in time and taking the most appropriate action as a prescription to avoid any type of risk. I hope you liked this article on the concept of descriptive and predictive analysis in Machine learning. Feel free to ask your valuable questions in the comments section below.