- Overview
- Motivation
- Technical Aspects
- EDA and Data Preprocessing
- Implementation of Regression Algorithm
- To Do
- Technology Used
- Team
- Credit
To predict a regression problem, like house sale price prediction, some regression algorithms are used to predict the best result. The model ./price_pred_model.pkl
takes a previous processed and scaled data set and predict the sale price of the house. This model can predict 85% accurately the rate of a house.
It's usually a big to predict or evalute the price of a house depending on some features manually. For this, using a proper machine learning algorithm, we can do this in no time.
- EDA and Preprocess of the dataset
In this notebook file, the data insights, trends are explored using
matplotlib
,pandas
,numpy
. Missing values and outliers have been treated. Using feature generation and transformation techniques, new effective features were added and unnecessary features were excluded.
- Implementation of Regression Model on the dataset
In this section,
i. The dataset was splitted into X(Observation Matrix) and y(Target Vector)
ii. Removal of Multicollinearity using the value of Variation Inflation Factor.
iii.Splitting intoX_train
,y_train
,X_test
,y_test
(training and test dataset).
iv.Implementation ofsklearn.linear_model.LinearRegression()
and checking for R2_score, Homoscedasticity and distribution of error.
v.Checking for importance of various feature given by the model.
Implementation ofsklearn.linear_model.Lasso()
andsklearn.tree.DecisionTreeRegressor
and the plot of the tree and the linear relation between the target values and the predicted values.
i. Implement a pipeline in this problem statement.
ii. Increament of the accuracy of the model
iii. Deployment.
This dataset is provided by Internshala Machine Learning Training.