Purpose and Technologies: The following program uses Python, Matplotlib, Numpy, Sympy, Pandas,and sci-kit to predict house prices in the city of Boston.
Procedure:
- Formulate a question
- Gather the data
- Clean the data
- Explore the possibilities and correlations
- Model the data graphically
- Evaluate the price within a function
Algorithm: I implemented a multivariable linear regression model to forecast house prices. This method, also known as multiple regression, is a statistical technique that uses several parameter variables to predict an outcome of a response variable. In this case, I extracted several variables from the dataset, including house prices, crime rates, age, proximity to water, tax, and more, to make my model more accurate. I picked out which variables had a low p-value corresponding to the initial price data, and also other conditions that had logical sense for this particular problem. I then applied my knowledge in distribution methods, standard deviation, MSE, RMSE etc ... to do a final calculation of what the house price would be.
Additionally, I take into account issues like Multicollinearity in linear regression, to make sure my program is as precise as possible.
Data Visualization Work:
Below is a model to demonstrate the different correlations between the variables in the dataset.
Another graphical representation of all the correlations between each variable; includes a regression fit
Statistical Analysis of the p values and coefficients for specific conditions:
