Empirical Analysis of Income Prediction Using Deep Learning Techniques

One of the main problems with determining income for employees is that there is no single formula that can be applied to all employees. Each employee's income is based on a variety of factors such as job position, experience, and special skills. Another problem is that the cost of living in different parts of the country can vary significantly, making it difficult to accurately assign a salary to different employees. The third issue is that there are variations in the types of benefits and perks that employers offer, which can make it difficult to accurately compare total compensation between different positions. Finally, there are often external factors that can affect the income of employees, such as the economic conditions of the area they are located in or the cost of health insurance. It is important to know what will be the income of employees to ensure that they are paid fair wages for their work and to ensure that total compensation is competitive with the market. Knowing the income of employees is also important to ensure that the employer can budget for salaries, benefits, and other costs associated with hiring and employing staff. To overcome the problems, we have made a model using the different models of LSTM. This paper presents a comparative study of various deep learning techniques for income prediction. Specifically, Long Short-Term Memory (LSTM), Stacked LSTM, Bidirectional LSTM, and Convolutional Long Short-Term Memory (Conv-LSTM) architectures are used to predict income. Simulations are processed on the Socio-Economic dataset having twenty-six years of monthly income data. This study provides insights into the effectiveness of deep learning techniques for predicting income.