REAL ESTATE PRICE ANALYSIS AND PREDICTION TOOLS FOR KATHMANDU VALLEY

REAL ESTATE PRICE ANALYSIS AND PREDICTION TOOLS FOR KATHMANDU VALLEY

December 17, 2019

ABSTRACT

The “Real Estate Price Analysis and Prediction Tools For Kathmandu Valley” is a market analysis tool that aims to predict the current commercial values of land based on series of available data. The project tends to meet the requirements of ‘Major Project’ for B.E. Fourth Year. Real Estate market is a large repository data. By integrating statistical analysis with data mining techniques those data can be analyzed and patterns can be generated from them. The project indexes to generate technical indicators from the collected data. Then regression analysis is created using those indicators as the inputs using regression analysis.

This project analyzes the collected real estate data and the current commercial values of the land of Kathmandu valley are predicted. This project can be helpful to determine the current commercial value of the land of Kathmandu valley. The overall functionalities of the project along with its specifications, design, methodology, result analysis are described in this report.

Key Words: Multiple Linear Regression, Lasso, PLR, WLR, KNN, ANN, Commercial Price.

INTRODUCTION

The real estate market is one of the most important markets in the modern economy because of the nature of the goods exchanged. Shelter is a fundamental human need and therefore there is a collective interest in pricing homes correctly. While there are other factors in play, an ill-informed party on either side of the transaction can be burned for non-trivial sums. Any insight into pricing lands properly and consistently would be of great interest to all parties involved.

A house is often the largest investment a person will make in their lifetime. The amount of money used to buy this house/Land is non-trivial and thus great care must be taken in not only choosing the right house, but making sure it’s priced appropriately. This knowledge is usual held solely by real estate agents. If we can capture this domain knowledge by using openly accessible data, then this knowledge suddenly becomes accessible to the average citizen who can thus make informed decisions without relying on an expert who unfortunately may not always be acting in their best interest.

On the side of the real estate agents, being able to accurately price a land through use of machine learning algorithm with available data will offer new insight in to what people are looking for in a land. They can thus spend more time focusing on creating successful pitches to sell the lands as well as make recommendations to clients.

Authors: Angad Gupta,Arun Kr Agrawal,Bikash Gupta, Shubham Kr Agrawal