Need a predictive model for predicting sales prices using python
Budget: $250 – $750 USD
Project Description:
I have a CSV file of sold properties in my city within the United States. There are 100s of thousands of records which can be selected from in my system.
I need an accurate predictive model to predict the following target variables:
1. Closed sales price of a house.
2. Closed sale price of a rental.
3. Closed sale price of a Hi-Rise Condominium
4. Closed sale price of a Townhouse
5. Number of days it takes to sell each of the above named categories.
Tasks:
Gathering the dataset:
Text mine property description for more data points.
Webscrape web pages for School data, crime rates, income levels.
Data preparation such as removing erroneous values, preparing flag variables, z-standardization, min-max standardization, prevent over-fitting using proper techniques (i.e. L-regulation, cross-validation, etc.), remove correlated variables, calculate accuracy of models to improve and obtain the best model, Explore different models to see which is best (i.e. Log Regression, MLP, KNN, Decision Trees, etc.)
Deliverables:
1. One predictive model for each target variable named above which has been tested and optimized to the best accuracy and not over-fitted.
2. All code provided so that I can use it in my applications. Request Jupyter Notebook format.
3. 90 days technical support for questions about the code or process.
I have a CSV file of sold properties in my city within the United States. There are 100s of thousands of records which can be selected from in my system.
I need an accurate predictive model to predict the following target variables:
1. Closed sales price of a house.
2. Closed sale price of a rental.
3. Closed sale price of a Hi-Rise Condominium
4. Closed sale price of a Townhouse
5. Number of days it takes to sell each of the above named categories.
Tasks:
Gathering the dataset:
Text mine property description for more data points.
Webscrape web pages for School data, crime rates, income levels.
Data preparation such as removing erroneous values, preparing flag variables, z-standardization, min-max standardization, prevent over-fitting using proper techniques (i.e. L-regulation, cross-validation, etc.), remove correlated variables, calculate accuracy of models to improve and obtain the best model, Explore different models to see which is best (i.e. Log Regression, MLP, KNN, Decision Trees, etc.)
Deliverables:
1. One predictive model for each target variable named above which has been tested and optimized to the best accuracy and not over-fitted.
2. All code provided so that I can use it in my applications. Request Jupyter Notebook format.
3. 90 days technical support for questions about the code or process.