HOUSE RENT
PREDICTION MODEL
AAKASH AGARWAL
ABHINAV ANAND
ANUGRAH SUNIT KISPOTTA
ARINDAM MANDAL
Data Introduction
● In this Dataset, we have information on almost 4700+
Houses/Apartments/Flats Available for Rent with different parameters
like-
• BHK • City
• Rent • Furnishing Status
• Size • Type of Tenant Preferred
• No. of Floors • No. of Bathrooms
• Area Type • Point of Contact
• Area Locality
Expected Outcome
● A thorough study of the data will aid in the discovery of valuable
insights that will be of enormous value to customers
● Predicting the rent of any house, based on the location, no. of
bathrooms, furnishing status, tenant, preferred size and on which the
floor of the building, etc
Frequency Distribution Based on BHK
• Mostly 2 BHK houses are rented the most, as it is mostly under budget for the people and can be rented
soon.
• 5 and 6 BHK houses are put less for rent because there is very less demand for such big spacious
houses in India.
Frequency Distribution Based on Bathroom
• House with 2 Bathrooms are most common for the houses put up on rent.
• Houses with 7 and 10 bathroom quite seems inappropriate and not much of use.
Frequency Distribution Based on City
• Mumbai, followed by Chennai and Hyderabad, has the most number of rented houses, it seems like there
is a very high demand considering the job corporates and other factors.
Frequency Distribution Based on Tenant Preference
• Mumbai, followed by Chennai and Hyderabad, has the most number of rented houses, it seems like there
is a very high demand considering the job corporates and other factors.
Correlation Matrix
● This determines how one
parameter is affected by a change
in another parameter.
● There is a strong correlation
observed between BHK and
Bathroom count.
However, rent and size don’t
seem to be much correlated
together.
BHK and Rent distribution
● The rent of the houses is
subsequently increasing with
increase in BHK.
● For 6 BHK, things doesn't seem
feasible and they are less in demand
so rent is very les
● The rent of the houses is
subsequently increasing with
increase in BHK.
● For 6 BHK, things doesn't seem
feasible and they are less in demand
so rent is very les
Area Type and Rent
• Houses with Carpet Area have
more rent as compared to Super
Area.
• Built Area are less in demand ,
so rent is kept less.
Boxenplot for City and Rent
• Seems Mumbai is having too much
demand for houses, so rent is also
very high. The immigrations of job
corporates to Mumbai is high. This
could be one of the reason of more
demand in rented houses.
• All other cities except Kolkata are
having fair equal amount of rent.
• Kolkata is not very much flourised
in terms of job sectors and lifestyle
as compared to other cities.
Moreover it is also less developed
compared to them. So demand is
less there, so rent is also less.
Furnishing
Status and Rent
• Unfurnished houses
for rent are having
less rent as
compared to others.
Tenant Preferred
and Rent
● Houses rented with no
tenant conditions are having
a bit more rent as compared
to others.
Linear Regression
Training data evaluation
of model result
MAE:197479077630.4328
MSE:9.768915536355845e+22
RMSE:312552644147.4435
Decision Tree
Training data evaluation
of model result
MAE: 0.25767891524169967
MSE: 0.4224998679675354
RMSE: 0.6499998984365578
Random Forest
Training data evaluation
of model result
MAE: 0.32968860404858025
MSE: 0.522145413635148
RMSE: 0.7225963005960853
Comparison of models
Model Performance Comparison: Evaluating
Regression Models
•Evaluation Metrics:
• Mean Absolute Error (MAE)
• Mean Squared Error (MSE)
• Root Mean Squared Error (RMSE)
•Metric Interpretation:
• Smaller metric values indicate
better model performance.
•Observation:
• Decision Tree Regressor
outperforms other models in all
metrics.
•Implication:
• Decision Tree Regressor is the
likely best-performing model
based on these metrics.
Who can use this model
Students who are newly
moving to these cities
post MBA
Based on the data Platform
can give suggestions for
best price range of a
The owners/brokers can set property
the right price for the
property
THANK
YOU
CREDITS: This presentation template was created by Slidesgo,
including icons by Flaticon, and infographics & images by Freepik
