Predictive modelling of business success using random forest, jrip and naïve Bayes algorithms.

Main Article Content

Emmanuel Olalekan Taiwo
Gbenga Oyewole Ogunsanwo
Olumuyiwa Bamidele Alaba

Abstract

The paper focused on the development of a predictive model for business success using machine learning algorithms. The model classifies and predicts business as either gain or loss. The historical dataset was collected for a period of ten (10) years (2012-2021) from Ogun State Chambers of Commerce, Industry Mines, and Agriculture Ltd/Gte. The dataset was further divided into 80% for training and 20% for testing. The metrics used for evaluation include: classification accuracy, execution time, error rate, ROC Area, mean absolute error (MAE), root mean squared error (RMSE) and confusion matrix. The dataset was used to formulate predictive models for business success using Random Forest, JRip and Naïve Bayes algorithms. Waikato Environment for Knowledge Analysis (WEKA) statistical tool was used to carry out the formulation and simulation of the predictive model. Results show Classification Accuracy (%) of 60.9, 63.9, 68.4. Execution Time (Seconds) of 0.68, 0.03, 0.1. Error Rate (%) of 39.0, 30.6, 31.6. ROC Area of 0.504, 0.509. 0.488. Mean Absolute Error (MAE) of 0.1715, 0.1683, 0.1717. Root Mean Squared Error (RMSE) of 0.3298, 0.2938, 0.2953 for Random Forest, JRip and Naïve Bayes algorithms respectively. The three models were compared and the best model in terms of accuracy and ROC Area was selected and validated. The study revealed that Naïve Bayes model has higher accuracy followed by JRip and Random Forest algorithms. The model is recommended for Business evaluation and any other machine learning algorithms can be used for business success predictive model.

Article Details

How to Cite
Taiwo, E. O., Ogunsanwo, G. O., & Alaba, O. B. (2023). Predictive modelling of business success using random forest, jrip and naïve Bayes algorithms . Faculty of Natural and Applied Sciences Journal of Scientific Innovations, 5(2), 30–37. Retrieved from https://fnasjournals.com/index.php/FNAS-JSI/article/view/217
Section
Articles