This is the first predictive analytics project as part of the Dicoding submission. The project involves building a machine learning model that can predict the rental prices of houses and apartments in India.
Background Housing, whether it's a house or an apartment, is a primary need for humans to live and shelter. The value of a residence depends on characteristics such as location, area, number of bedrooms, number of bathrooms, furnishings, and other features.
The price of each home is determined by its value. However, it is challenging to manually predict the rental price accurately. To reduce uncertainty, rental companies can develop predictive systems that determine appropriate rental prices for homes with specific characteristics.
The goal of this research is to build a machine learning model that can predict rental prices aligned with the market. This prediction will serve as a reference for companies to offer rental prices that can generate profit.
Analysis of House Price Prediction Based on Specifications Using Multiple Linear Regression
This project is designed for companies with the following business characteristics:
The company owns or purchases homes and apartments and then rents them to consumers. The company offers consulting services on house and apartment rental prices to consumers.
Which features influence house or apartment rental prices the most? How can data be processed to be effectively trained by models? What is the market rental price for a house with specific characteristics?
Identify the features that most affect house or apartment rental prices. Prepare the data for model training. Build a machine learning model that can accurately predict rental prices based on specific characteristics.
Analyze data through univariate and multivariate analysis. Data visualization helps in understanding feature correlations and detecting outliers. Prepare the data for model training. Perform hyperparameter tuning using grid search and build a regression model to predict continuous values. The algorithms used in this project include K-Nearest Neighbour, Random Forest, and AdaBoost.