live demo https://business-force.streamlit.app/
- in this video, we are going to make a machine learning web project driven by a probabilistic model that is multivariable linear regression.
- there are three variables the purpose is to check how far the linear relationship is between these variables, where x has two features as independent variables and y has one feature as dependent variable. This is a classification problem using multiple regression analysis for the data in MySQL. finally visualizing the measure of variations and lines of the best fit
- we will fetch data from MySQL database. then using Python, we will load data and then we will train the model, and visualize an output
- Whenever you open the website or refresh the page, python loads the data from MySQL and retrains the data every time you load the page. and this makes the model learn more as the data enters the database
- First of all, we will check if there is a relationship between the variables and if there is, is it positive or negative, that is the slope in short. this is called regression. then we will look at how strong the relationship is. this is called the correlation coefficient
- then we will look at how much percentage change in X leads to change in Y, this is called the coefficient of determination
- load data from MySQL using python
- basic data exploration
- data modeling & feature selection
- model training & model fitting
- prediction
- X-features to be standard normal distribution (z score)
- finding regression equation line
- finding the coefficient of correlation (r)
- finding the coefficient of determination (R2)
- finding the adjusted coefficient of correlation
- the regression equation and line of best fit
- mean squared error
- mean absolute error
- root mean squared error
- normal distribution curve
