Impact of Predictive Analytics and Data mining in Business

Today to survive in this knowledge-based economy business needs the ability to transform data into knowledge. Digital revolution provided us with convenience and ease of access, storage and distribution of vast amount of data. A support mechanism for maintaining such a huge data is required; this mechanism not only includes men and machine but also those models which provide them enough support to take a decision, anticipating business scenarios.

In recent years data mining techniques have been widely used for discovering interesting, non-obvious relationships hidden in a data to predict future trends and forecast possible opportunities.

Predictive analytics is the branch of data mining that use data, statistics and machine-learning techniques for prediction of future probabilities and trends to produce new sagacity that leads to best decision for a given situation. The pivotal component i.e. predictant and predictor is evaluated to predict future trends. Multiple predictors can be combined to generate a predictive model for forecasting future probabilities. When used together, predictive analytics and data mining can make business more efficient. Data mining is considered as analytic toolset that automatically retrieve hidden patterns and predictive analytics is a guided discipline to make forward-looking predictions.

Some of the popular techniques are

1.      Linear regression: It is a method to estimate the unknown effect of changing one variable over another.

2.      Time series forecasting: This technique uses model to predict the future values based on previously recorded values. Examples include stock prices and weather forecasting.

3.      Bayesian analysis: It captures the concepts used in probability forecasting. It is a statistical procedure which estimates parameters of an underlying distribution based on the observed distribution.

4.      Regression analysis: It is a statistical tool for the investigation of relationships between variables. It seeks to ascertain the effect of one variable upon another-for example the effect of a price increase upon demand.

5.      Classification: It uses attributes in data to assign an object to a predefined class or predict the value of a variable of interest. Examples include credit risk analysis, likelihood to purchase.

6.      Clustering separates data into homogeneous subgroups based on the properties of the data. Clusters are formed with high intra cluster similarity. An example is customer demographic segmentation.

Application of Predictive Analytic

Fraud detection and security – Predictive analytic detect fraudulent activity before they occur and minimize the financial losses. By combining multiple detection methods we can get greater accuracy and better predictive performance.

Marketing – predictive analytics is used to understand customer buying patterns, their preferences, as well as to promote sales. It also helps in identifying potential buyers and maximize their marketing spending.

Operations Management – Nowadays most of the companies are using predictive models to forecast inventory and manage factory resources. For example Hotels try to predict the number of guests on any given occassion to adjust prices to maximize occupancy and increase revenue. Predictive analytics enables businesses to work more efficiently and effectively. 

Health care providers: Predict the effectiveness of new procedures, disease diagnostics i.e. to examine severity in certain conditions like heart disease, cancer, etc and improve ser­vices by providing safe and effective patient care.

Mr. Varun Sapra

Assistant Professor
Deptt. of Information Technology