Machine Learning for cost prediction
Introduction
Predictive analytics brings together a variety of statistical modeling, machine learning, and data mining techniques that analyze actual current and historical data to make predictions about the future or unknown events.[1][2].
In business, predictive models extract patterns from historical and transactional data to identify risks and opportunities. Predictive models identify relationships between different factors that allow the assessment of associated risks or probabilities based on a set of conditions, thus guiding the decision-maker during the organization's operations.[3].
The functional effect intended by these technical initiatives is that predictive analysis provides a score (probability) for each subject (customer, employee, patient, product, vehicle, component, machine and other unit in the organization) in order to determine, inform or influence processes in the organization in which a large number of subjects participate, as occurs in marketing, credit risk assessment, fraud detection, manufacturing, health and government operations such as law enforcement.
Predictive analytics is being used in casinos,[4] actuarial science,[5] e-commerce,[6] finance,[7] government,[8] pharmaceutical industry,[9] marketing,[10] retail,[11] insurance company,[12] telecommunications,[13] healthcare,[14] travel[15] and other fields.
One of the best-known applications is the credit score used in financial services. Scoring models process a client's credit history, loan applications, client data, etc., in order to order and classify subjects by their probability of being able to make future payments on time.
Definition
Predictive analysis is an area of data mining that aims to extract knowledge that allows it to predict trends and behavioral patterns. Often an unknown circumstance of interest will occur in the future, but predictive analytics can equally be applied to the unknown whether in the past, present or future. For example, identifying suspects after a crime or credit card fraud has occurred.[16] The fundamental thing about predictive analysis is identifying relationships between the explanatory variables and the predictive variables of the past so that it can be scaled to what is about to happen. It is important to note, in any case, that the reliability and usability of the results will depend greatly on the level of data analysis and the quality of the hypotheses.