Improving Organizational Decision-Making Using Random Forests and Big Data Analytics

This research aims at exploring how Random Forests can be used and Big Data Analytics to improve on decision making in organizations. The research is based on a large data set of 200 gb that contains information on the customer, employer’s satisfaction, operational efficiency data and market data. Random forest model has shown changes in decision making where they made lot of improvements in the accuracy of decisions in different fields. The accuracy for customer churn prediction increased from 72.5% (traditional methods) to 90.1% using Random Forests, a 17.6% improvement. Similarly, the accuracy of fraud detection improved from 78.4% to 92.3%, a gain of 13.9%. Feature importance analysis identified Customer Lifetime Value (importance score: 0.152) and Transaction Frequency (importance score: 0.140) as critical factors driving decisions. Furthermore, Random Forests have been incorporated with Big Data Analytics, which cut the processing time to up to 79%. It gives only 2% when analyzing 1 TB data, which also confirm the proposed approach is scalable and effective. As is seen from the study, the use of data-driven models can go a long way in enhancing decision making and providing strategic solutions for bettering of most organizational operations i. e. marketing, operations, and human resource. More specifically, the results presented in this paper point at the potential benefits that organizations may derive from the adoption of machine learning models and big data techniques: better, quicker and more effective decisions.