Big Data Application Areas

The clustering model belongs to the splash-reading mining model, which takes user's attribute, behavior, consumption and other characteristic data as input, and automatically clustering the users into several classes, which is often used to excavate potential target customers and can be used in big data marketing tools, CRM tools and fraud prevention solutions.

The classification prediction model analyzes the experience of learning historical data and predicts the future development direction of data. Model output is a discrete data or category called a classification model, and model output is a model of numeric type data called a numerical prediction model. The classification model constructs the classifier according to the class number attribute of the training data set, learns the classification rules of the existing classification data, and finally spares the new data in the classification. The numerical prediction model fits the training data according to the data input, and finally establishes the continuity numerical function. Typical applications of categorical predictive models include fraud detection, market positioning, performance prediction, medical diagnosis, and price forecasting.

Association rule Mining Model, which is used to discover interesting connections hidden in a large data set. Association rule mining is divided into "frequent itemsets generation" and "Association Rule Generation" two main tasks. The purpose of frequent itemsets generation is to discover all itemsets that meet the minimum support threshold, which are called frequent itemsets. The "Generation of association rules" goal is to extract all high-confidence rules from the frequent itemsets found in the previous step, which are called strong rules. Typical application of association rule Mining Model product correlation recommendation and precise marketing.

Big Data Application Areas