Predicting House Price in India Using Linear Regression Machine Learning Algorithms

Abstract

The price of house is increased every year according to the location and it indicates the current economic situation so there is a need for a system to predict house sales in the future for both buyer and the seller. Here we use dataset of India across different cities and having more than 68,613 entries of train data and 28000 test data of housing sales in whole, India. This analysis includes the effect of markdowns on the sales and the extent of effects on the sales by size, price, area etc. has been analyzed using different machine learning algorithms. Estimating home sales can help the developer determine the selling price of the home and the best time for the buyer to purchase the home. The output values of the algorithms are estimated based on the input characteristics from the data presented in the system and the analysis is a process. Physical conditions, concept and location are the three factors that determine the selling price of a home.An exact forecast of forthcoming development market interest, particularly the private market, is central imperative to strategy producers, as it could assist with forming procedures to develop/balance out the economy and fulfill the social requirements. In spite of that, a sensible forecast of future private interest is never a simple assignment, as it is represented by various social and monetary elements. In this paper, four proactive factor models are created and thought about for straightforwardly anticipating India private area a wide range of interest. In giving a scope of potential estimates, the model additionally gives a chance to the chief to practice judgment in choosing the most fitting figures. We depict the continuous internet based customer conduct expectation framework which predicts the clients buying aim when the client visits the site. To achieve the assignment, we rely upon meeting and guest data and we utilize gullible Bayes classifier, Logistic Regression, Neural Network, GBC, choice tree and arbitrary timberland research the dataset for furcating the buyer choice. Also, we use oversampling to work on the exhibition and the adaptability of the classifier. The outcomes show that irregular backwoods creates the higher precision and utilized different Algorithms. We study about the algorithm linear Regression taking a data set.