Predictive Analytics have been widely used in Payment and Banking applications for transaction processing. Companies are using various algorithms to determine the credit limit and spending capacity for an individual. With the rise in data collection and data analytics advancements, this process needs to be more inclusive for people with mental disorders such as Bipolar Disorder, which can help them get control over their finances. In order to conduct this research analysis, a real-world anonymized dataset is used with few synthetic fields. To improve the effectiveness of different algorithms, concept of Dimensionality Reduction is utilized. To determine the accuracy, various Supervised Machine Learning classification algorithms have been implemented. Findings revealed that Random Forest was most accurate with an accuracy of 0.951 using the RFCEV method, while the Artificial Neural Network gave the best results with PCA of 0.913. By deploying the solution on public cloud network, it can be used as Mobile first application to help users with mental disorders monitor, analyze and control their spending in real-time. This will help in making banking and payment industry more inclusive and improve accessibility.