https://raw.githubusercontent.com/Masterx-AI/Project_Housing_Price_Prediction_/main/hs.jpg" alt="">

Description:

A simple yet challenging project, to predict the housing price based on certain factors like house area, bedrooms, furnished, nearness to mainroad, etc. The dataset is small yet, it's complexity arises due to the fact that it has strong multicollinearity. Can you overcome these obstacles & build a decent predictive model?

Acknowledgement:

Harrison, D. and Rubinfeld, D.L. (1978) Hedonic prices and the demand for clean air. J. Environ. Economics and Management 5, 81–102. Belsley D.A., Kuh, E. and Welsch, R.E. (1980) Regression Diagnostics. Identifying Influential Data and Sources of Collinearity. New York: Wiley.

Objective:

• Understand the Dataset & cleanup (if required).
• Build Regression models to predict the sales w.r.t a single & multiple feature.
• Also evaluate the models & compare thier respective scores like R2, RMSE, etc.

The number of U.S. home sales in the United States declined in 2024, after soaring in 2021. A total of four million transactions of existing homes, including single-family, condo, and co-ops, were completed in 2024, down from 6.12 million in 2021. According to the forecast, the housing market is forecast to head for recovery in 2025, despite transaction volumes expected to remain below the long-term average. Why have home sales declined? The housing boom during the coronavirus pandemic has demonstrated that being a homeowner is still an integral part of the American dream. Nevertheless, sentiment declined in the second half of 2022 and Americans across all generations agreed that the time was not right to buy a home. A combination of factors has led to house prices rocketing and making homeownership unaffordable for the average buyer. A survey among owners and renters found that the high home prices and unfavorable economic conditions were the two main barriers to making a home purchase. People who would like to purchase their own home need to save up a deposit, have a good credit score, and a steady and sufficient income to be approved for a mortgage. In 2022, mortgage rates experienced the most aggressive increase in history, making the total cost of homeownership substantially higher. Are U.S. home prices expected to fall? The median sales price of existing homes stood at 413,000 U.S. dollars in 2024 and was forecast to increase slightly until 2026. The development of the S&P/Case Shiller U.S. National Home Price Index shows that home prices experienced seven consecutive months of decline between June 2022 and January 2023, but this trend reversed in the following months. Despite mild fluctuations throughout the year, home prices in many metros are forecast to continue to grow, albeit at a much slower rate.

This dataset contains data on all Real Property parcels that have sold since 2013 in Allegheny County, PA.

Before doing any market analysis on property sales, check the sales validation codes. Many property "sales" are not considered a valid representation of the true market value of the property. For example, when multiple lots are together on one deed with one price they are generally coded as invalid ("H") because the sale price for each parcel ID number indicates the total price paid for a group of parcels, not just for one parcel. See the Sales Validation Codes Dictionary for a complete explanation of valid and invalid sale codes.

Sales Transactions Disclaimer: Sales information is provided from the Allegheny County Department of Administrative Services, Real Estate Division. Content and validation codes are subject to change. Please review the Data Dictionary for details on included fields before each use. Property owners are not required by law to record a deed at the time of sale. Consequently the assessment system may not contain a complete sales history for every property and every sale. You may do a deed search at http://www.alleghenycounty.us/re/index.aspx directly for the most updated information. Note: Ordinance 3478-07 prohibits public access to search assessment records by owner name. It was signed by the Chief Executive in 2007.

1. Customers File (customers.csv)

• Description: This file contains information about clients involved in real estate transactions. It includes personal details such as name, surname, birth date, gender, and country, along with transaction-specific information like the purpose of the deal and the satisfaction level.
• Key Columns:
  • customerid: Unique identifier for the customer.
  • entity: Type of client, whether an individual or a company.
  • name and surname: First and last name of the customer.
  • birth_date: Customer's date of birth.
  • sex: Gender of the customer (Male/Female).
  • country and state: The country and state the customer is associated with.
  • purpose: Purpose of the transaction (e.g., Home purchase or Investment).
  • deal_satisfaction: Customer's satisfaction level with the transaction, ranging from 1 to 5.
  • mortgage: Indicates whether the transaction involved a mortgage (Yes/No).
  • source: How the customer was acquired (e.g., Website or Agency).

2. Properties File (properties.csv)

• Description: This file contains information about the properties sold, including building details, property type, area, price, and sale status.
• Key Columns:
  • id: Unique identifier for the property.
  • building: Number of the building where the property is located.
  • date_sale: The date when the property was sold.
  • type: Type of property (e.g., Apartment).
  • property#: The property number within the building.
  • area: Area of the property in square feet.
  • price: Sale price of the property.
  • status: Status of the sale (e.g., Sold).
  • customerid: The unique identifier of the customer associated with the property.

Suggested Analysis and Tasks

1 Customer Insights: - Customer Segmentation: Group customers based on demographics, purpose, or deal satisfaction to understand different customer profiles. - Satisfaction Analysis: Investigate what factors (e.g., property price, area, or mortgage involvement) influence customer satisfaction levels. - Source Effectiveness: Analyze which acquisition sources (e.g., website or agency) yield the highest deal satisfaction.

2 Property Market Analysis: - Price Trends: Analyze how property prices vary over time or by location to identify market trends. - Demand Analysis: Determine which types of properties (e.g., apartments vs. houses) are most popular based on sales data. - Area vs. Price: Explore the relationship between property area and price to develop pricing models or evaluate property value.

3 Predictive Modeling: - Price Prediction: Build models to predict property prices based on features like area, type, and location. - Satisfaction Prediction: Create models to predict customer satisfaction using transaction details and demographics. - Likelihood of Sale: Develop a model to predict the likelihood of a property being sold based on its attributes and market conditions.

4 Geographical Analysis: - Heatmaps: Create heatmaps to visualize property sales and identify high-demand areas. - Country and State Trends: Examine how real estate trends differ between countries and states.

5 Mortgage Impact Study: - Mortgage vs. Non-Mortgage Analysis: Compare transactions that involved a mortgage to those that didn’t to study the impact on price, satisfaction, and deal closure speed.

6 Time Series Analysis: - Sales Over Time: Analyze property sales over different periods to identify seasonal trends or patterns. - Customer Birth Date Analysis: Study any correlations between customers’ birth years and their purchasing behavior.

Graph and download economic data for Median Sales Price of Houses Sold for the United States (MSPUS) from Q1 1963 to Q2 2025 about sales, median, housing, and USA.

Graph and download economic data for Average Sales Price of Houses Sold for the United States (ASPUS) from Q1 1963 to Q2 2025 about sales, housing, and USA.

The U.S. housing market continues to evolve, with the median price for existing homes forecast to fall to ******* U.S. dollars by 2027. This projection comes after a period of significant growth and recent fluctuations, reflecting the complex interplay of economic factors affecting the real estate sector. The rising costs have not only impacted home prices but also down payments, with the median down payment more than doubling since 2012. Regional variations in housing costs Home prices and down payments vary dramatically across the United States. While the national median down payment stood at approximately ****** U.S. dollars in early 2024, homebuyers in states like California, Massachusetts, and Hawaii faced down payments exceeding ****** U.S. dollars. This disparity highlights the challenges of homeownership in high-cost markets and underscores the importance of location in determining housing affordability. Market dynamics and future outlook The housing market has shown signs of cooling after years of rapid growth, with a modest price increase of *** percent in 2024. This slowdown can be attributed in part to rising mortgage rates, which have tempered demand. Despite these challenges, most states continued to see year-over-year price growth in 2025, with Rhode Island and West Virginia leading the packby home appreciation. As the market adjusts to new economic realities, potential homebuyers and investors alike will be watching closely for signs of stabilization or renewed growth in the coming years.

Purpose and brief description The house price index measures the inflation in the residential property market. The house price index reflects price developments for all residential properties purchased by households (apartments, terraced houses, detached houses), regardless of whether they are new or existing. Only market prices are taken into account, so self-build homes are excluded. The price of the land is included in the price of the properties. Population Real estate transactions involving residential properties Periodicity Quarterly. Release calendar Results available 3 months after the reference period Definitions House price index: The house price index measures changes in the prices of new or existing dwellings, regardless of their use or previous owner. Inflation - house price index: Inflation is defined as the ratio between the value of a given quarter and that of the same quarter of the previous year. Weighting - house price index: Weighting based on the national accounts (gross fixed capital formation in housing) and the total number of real estate transactions involving residential properties. Type of dwelling according to the classification set out in Regulation (EU) No 93/2013 on housing price indices. Technical information The house price index measures the price evolution of real estate prices on the market of private property. The index follows price changes of new or existing residential real estate purchased by households, irrespective of their purpose (letting or owner-occupying). Only market prices are taken into account. Houses built by their owners are therefore not included. The price of the building plot is included in the house price. The house price index is based on real estate transaction data from the General Administration of the Patrimonial Documentation of the FPS Finances. The prices used are those included in the deeds of sale. Given the time between the date on which the preliminary sales agreement is signed and the date on which the deed is executed (between three and four months), this index measures the price evolution with a delay compared to the actual date on which the sales price is set. This delay is inherent to the data source. The house price index is calculated by the European Union Member States, Norway and Iceland. Eurostat calculates the index for the Euro area (as well as for the European Union as a whole) using the harmonised indices of the Member States. Given the role of the housing market in the economic and financial crisis of 2008, the house price index was included in the indicators used in the procedure to prevent and correct macroeconomic imbalances in the European Union. The house price index is calculated under the European Regulation 2016/792 on harmonised indices of consumer prices and the house price index and 2023/1470 laying down the methodological and technical specifications as regards the house price index and the owner-occupied housing price index. Data are available from 2005 onward for Belgium as well as for the European Union and the majority of European countries. The house price index can be broken down by new houses and existing houses. The weights of these two items in the overall index are determined by the gross fixed capital formation in houses (for the new houses) and the total value of transactions of the previous year (for the existing houses). Until 2013, the house price index of new houses was roughly estimated based on the output price index in the construction sector. Since 2014, it is also based on real estate transaction data. House price index for existing houses is available per region since 2010. The data have therefore been completely reviewed when the results for the fourth quarter of 2023 were published in March 2024. Since the houses that are put up for sale differ from one quarter to another, the changes in characteristics are processed with hedonic regression models to eliminate price fluctuations due to changes in characteristics of the properties sold. These models aim to estimate the theoretical price based on the characteristics and location of the houses sold. The index is then calculated based on changes in the average prices observed and adjusted by a factor depending on the differences in quality observed between dwellings sold during the different periods.

The median sales price of new homes sold in the United States increased steadily from 1965 to 2022, followed by two years of decline. In 2024, a newly built home cost approximately ******* U.S. dollars. That was a decline from the peak price of 434,500 U.S. dollars in 2022. Prices varied greatly across different regions in the country, with the most expensive housing found in the Northeast region.

The average sales price of new homes in the United States experienced a slight decrease in 2024, dropping to 512,2000 U.S. dollars from the peak of 521,500 U.S. dollars in 2022. This decline came after years of substantial price increases, with the average price surpassing 400,000 U.S. dollars for the first time in 2021. The recent cooling in the housing market reflects broader economic trends and changing consumer sentiment towards homeownership. Factors influencing home prices and affordability The rapid rise in home prices over the past few years has been driven by several factors, including historically low mortgage rates and increased demand during the COVID-19 pandemic. However, the market has since slowed down, with the number of home sales declining by over two million between 2021 and 2023. This decline can be attributed to rising mortgage rates and decreased affordability. The Housing Affordability Index hit a record low of 98.1 in 2023, indicating that the median-income family could no longer afford a median-priced home. Future outlook for the housing market Despite the recent cooling, experts forecast a potential recovery in the coming years. The Freddie Mac House Price Index showed a growth of 6.5 percent in 2023, which is still above the long-term average of 4.4 percent since 1990. However, homebuyer sentiment remains low across all age groups, with people aged 45 to 64 expressing the most pessimistic outlook. The median sales price of existing homes is expected to increase slightly until 2025, suggesting that affordability challenges may persist in the near future.

View monthly updates and historical trends for US Existing Home Median Sales Price. from United States. Source: National Association of Realtors. Track ec…

Single Family Home Prices in the United States increased to 415200 USD in October from 412300 USD in September of 2025. This dataset provides - United States Existing Single Family Home Prices- actual values, historical data, forecast, chart, statistics, economic calendar and news.

The distribution of the results of the sale prices of the real estate of the courts produced from 2006 to 2021 in France, will allow you to know or visualise the property value of a property sold at public auction in France, within a radius of an address, in accordance with the DVF file (Demands de Valeurs Foncières) and the regulations of the Etalab. The use of these files does not allow the exact identification of the assets of the persons concerned, which remain anonymous.

The reuse of these datasets below is free of charge, and allows to view, download and integrate the data of these results “**under condition**”:by mentioning the authorship of the source of this data by indicating one or more links to the platform of Les Encheres-Publiques.com. If you have any questions, do not hesitate to contact us at data@encherespubliques.com.

In the file “results-sales-2006-2021.csv”, you will find for each property what characterises it, according to:

— The date of sale, — The name of the organiser of the sale, — The category of the property, — The name of the road, the city and its postal code, — The surface of the property in m², — The link to the description of the lot, — And the result of its sale.

The file “court-judicial.csv”, will allow you to link the results of the sales prices to the contact details of the organisers by:

— The name of the judicial court, — The address of the place of sale, — And the link to its current sales,

Update files

The files made available are updated annually. Each update removes and replaces all previously released files. The last update dates from 28 September 2022.

To find out more: https://www.encheres-publiques.com/ventes/immobilier

Data on the number of residential properties sold, sale price and number of buyers by sale type, property type and period of construction.

Extract detailed property data points — address, URL, prices, floor space, overview, parking, agents, and more — from any real estate listings. The Rankings data contains the ranking of properties as they come in the SERPs of different property listing sites. Furthermore, with our real estate agents' data, you can directly get in touch with the real estate agents/brokers via email or phone numbers.

A. Usecase/Applications possible with the data:

1. Property pricing - accurate property data for real estate valuation. Gather information about properties and their valuations from Federal, State, or County level websites. Monitor the real estate market across the country and decide the best time to buy or sell based on data

2. Secure your real estate investment - Monitor foreclosures and auctions to identify investment opportunities. Identify areas within special economic and opportunity zones such as QOZs - cross-map that with commercial or residential listings to identify leads. Ensure the safety of your investments, property, and personnel by analyzing crime data prior to investing.

3. Identify hot, emerging markets - Gather data about rent, demographic, and population data to expand retail and e-commerce businesses. Helps you drive better investment decisions.

4. Profile a building’s retrofit history - a building permit is required before the start of any construction activity of a building, such as changing the building structure, remodeling, or installing new equipment. Moreover, many large cities provide public datasets of building permits in history. Use building permits to profile a city’s building retrofit history.

5. Study market changes - New construction data helps measure and evaluate the size, composition, and changes occurring within the housing and construction sectors.

6. Finding leads - Property records can reveal a wealth of information, such as how long an owner has currently lived in a home. US Census Bureau data and City-Data.com provide profiles of towns and city neighborhoods as well as demographic statistics. This data is available for free and can help agents increase their expertise in their communities and get a feel for the local market.

7. Searching for Targeted Leads - Focusing on small, niche areas of the real estate market can sometimes be the most efficient method of finding leads. For example, targeting high-end home sellers may take longer to develop a lead, but the payoff could be greater. Or, you may have a special interest or background in a certain type of home that would improve your chances of connecting with potential sellers. In these cases, focused data searches may help you find the best leads and develop relationships with future sellers.

How does it work?

• Analyze sample data
• Customize parameters to suit your needs
• Add to your projects
• Contact support for further customization

House Price Prediction Dataset.

The dataset contains 2000 rows of house-related data, representing various features that could influence house prices. Below, we discuss key aspects of the dataset, which include its structure, the choice of features, and potential use cases for analysis.

1. Dataset Features

The dataset is designed to capture essential attributes for predicting house prices, including:

Area: Square footage of the house, which is generally one of the most important predictors of price. Bedrooms & Bathrooms: The number of rooms in a house significantly affects its value. Homes with more rooms tend to be priced higher. Floors: The number of floors in a house could indicate a larger, more luxurious home, potentially raising its price. Year Built: The age of the house can affect its condition and value. Newly built houses are generally more expensive than older ones. Location: Houses in desirable locations such as downtown or urban areas tend to be priced higher than those in suburban or rural areas. Condition: The current condition of the house is critical, as well-maintained houses (in 'Excellent' or 'Good' condition) will attract higher prices compared to houses in 'Fair' or 'Poor' condition. Garage: Availability of a garage can increase the price due to added convenience and space. Price: The target variable, representing the sale price of the house, used to train machine learning models to predict house prices based on the other features.

2. Feature Distributions

Area Distribution: The area of the houses in the dataset ranges from 500 to 5000 square feet, which allows analysis across different types of homes, from smaller apartments to larger luxury houses. Bedrooms and Bathrooms: The number of bedrooms varies from 1 to 5, and bathrooms from 1 to 4. This variance enables analysis of homes with different sizes and layouts. Floors: Houses in the dataset have between 1 and 3 floors. This feature could be useful for identifying the influence of multi-level homes on house prices. Year Built: The dataset contains houses built from 1900 to 2023, giving a wide range of house ages to analyze the effects of new vs. older construction. Location: There is a mix of urban, suburban, downtown, and rural locations. Urban and downtown homes may command higher prices due to proximity to amenities. Condition: Houses are labeled as 'Excellent', 'Good', 'Fair', or 'Poor'. This feature helps model the price differences based on the current state of the house. Price Distribution: Prices range between $50,000 and $1,000,000, offering a broad spectrum of property values. This range makes the dataset appropriate for predicting a wide variety of housing prices, from affordable homes to luxury properties.

3. Correlation Between Features

A key area of interest is the relationship between various features and house price: Area and Price: Typically, a strong positive correlation is expected between the size of the house (Area) and its price. Larger homes are likely to be more expensive. Location and Price: Location is another major factor. Houses in urban or downtown areas may show a higher price on average compared to suburban and rural locations. Condition and Price: The condition of the house should show a positive correlation with price. Houses in better condition should be priced higher, as they require less maintenance and repair. Year Built and Price: Newer houses might command a higher price due to better construction standards, modern amenities, and less wear-and-tear, but some older homes in good condition may retain historical value. Garage and Price: A house with a garage may be more expensive than one without, as it provides extra storage or parking space.

4. Potential Use Cases

The dataset is well-suited for various machine learning and data analysis applications, including:

House Price Prediction: Using regression techniques, this dataset can be used to build a model to predict house prices based on the available features. Feature Importance Analysis: By using techniques such as feature importance ranking, data scientists can determine which features (e.g., location, area, or condition) have the greatest impact on house prices. Clustering: Clustering techniques like k-means could help identify patterns in the data, such as grouping houses into segments based on their characteristics (e.g., luxury homes, affordable homes). Market Segmentation: The dataset can be used to perform segmentation by location, price range, or house type to analyze trends in specific sub-markets, like luxury vs. affordable housing. Time-Based Analysis: By studying how house prices vary with the year built or the age of the house, analysts can derive insights into the trends of older vs. newer homes.

5. Limitations and ...

Update 10/31/2023: Sales are no longer filtered out of this data set based on deed type, sale price, or recency of sale for a given PIN with the same price. If users wish to recreate the former filtering schema they should set sale_filter_same_sale_within_365, sale_filter_less_than_10k, and sale_filter_deed_type to False.

Parcel sales for real property in Cook County, from 1999 to present. The Assessor's Office uses this data in its modeling to estimate the fair market value of unsold properties.

When working with Parcel Index Numbers (PINs) make sure to zero-pad them to 14 digits. Some datasets may lose leading zeros for PINs when downloaded.

Sale document numbers correspond to those of the Cook County Clerk, and can be used on the Clerk's website to find more information about each sale.

NOTE: These sales are filtered, but likely include non-arms-length transactions - sales less than $10,000 along with quit claims, executor deeds, beneficial interests are excluded. While the Data Department will upload what it has access to monthly, sales are reported on a lag, with many records not populating until months after their official recording date.

Current property class codes, their levels of assessment, and descriptions can be found on the Assessor's website. Note that class codes details can change across time.

For more information on the sourcing of attached data and the preparation of this dataset, see the Assessor's Standard Operating Procedures for Open Data on GitHub.

Read about the Assessor's 2025 Open Data Refresh.

Average House Prices in the United States increased to 534100 USD in August from 478200 USD in July of 2025. This dataset includes a chart with historical data for the United States New Home Average Sales Price.