Abstract:
Dhaka is a densely populated megacity in Bangladesh and day by day population of Dhaka city is increasing. People are frequently migrating to Dhaka. The main driving forces behind migration to Dhaka City are poverty, job searching, and family pressure. As a result, Dhaka is currently overpopulated. So, with all available houses, it is tough to accommodate this overpopulated Dhaka city. For this reason, rental house is one of the most severe issues in this city. House rental prices are having an effect by various factors. So, it is very important to determine the house rent price. The main goal of our work is to analyze the different features of a house and predict the rental price of a house based on multiple factors using machine learning algorithms. So, this report’s effort is to build a model that can predict house rent in Dhaka city. The area of Dhaka city namely Bashundhara, Gulshan, Mohammadpur, Uttara, Farmgate, Dhanmondi, Baridhara, Mirpur, Nikunja, Khilgaon, etc. Multiple factors including geographical location, house size, number of bedrooms, and number of bathrooms are considered. This work uses various machine-learning regression techniques to predict house rent and compare the accuracy of each algorithm. The following selected algorithms are - Linear Regression, Ridge Regression, Bayesian Regression, and Lasso Regression. All the proposed model gives almost similar accuracy. And our proposed model Ridge regression gives the highest accuracy with an accuracy of 91.54% on the other hand, Linear Regression, Bayesian Regression, and Lasso Regression give an accuracy of 91.49%, 91.52%, and 91.49%, respectively
