House_Price_Prediction

House Price Prediction using RandomForest and Linear Regression

Here, we are provided with a dataset with house features and its price. We are using 2 models to predict the price and confirm the best model

Why Random Forest

1. Random forest is an ensemble of decision trees.
2. This is to say that many trees, constructed in a certain “random” way form a Random Forest.
3. Each tree is created from a different sample of rows and at each node, a different sample of features is selected for splitting.
4. Each of the trees makes its own individual prediction.
5. These predictions are then averaged to produce a single result, hence improving its accuracy and reducing overfitting Here, practically we can see that the MSE(Mean Square Error) for Random forest is low compared to Linear Regression.

Steps Followed:

1. Load and View the dataset
2. Perform EDA(statistics, missing values, duplicates, covariance and correlation, relationship between variables)
3. Choose the regression model (Random Forest and Linear Regression)
4. Build a pipeline for automating the flow of data
5. Save the model using pickle

Streamlit

1. I have used streamlite to deploy the model.
2. Follow this tutorial to do the same in colab- https://faun.dev/c/stories/neji_14/how-to-create-and-launch-a-streamlit-app-directly-from-google-colab/