BookingGauger: Automated Machine Learning API for predicting number of accommodation days an online visitor will book.
This project builds a data processing pipeline and modeling with Automated Machine Learning (AutoML) to request prediction of number of days to be booked for accommodation by a user.
The aim of this project is to predict the number of days that customers are likely to book a room for an accommodation based on user bahaviour. The end-user is an accommodation provider who sought to obtain an intelligent tool that can enable the prediction of number of days that an online vistor will book a room for accommodation based on a number of features.
The dataset had a number of variables used as predictors for predicting number of accommodations booked as the target variable. These includes the following;
Number of sessions : This describes the number of sessions a customer made on the booking site.
City : This is the city from which a customer is accessing the booking site from
Country : This is the country from which the user is accessing the booking site. During the selection of various variables, you do not have the burden to decide this as reference is automatically made from the city selected.
Device Class : This is the type of device used to access the booking site. It has the values desktop, phone or tablet
Instant Booking : The is a feature on a booking site. Whether or not this feature was used by a customer is included in predicting the number of day to be booked
User Verification Status : Whether or not a customer who visited the site has been verified is included in predicting number of days to be booked.
__ Number of accommodation days to be booked__
Automated machine learning (AutoML) was employed to deliver a high accuracy optimized prediction model. The model is used to create an API that receives request, makes and send prediction as response to an app.
A reusable package that is truely plug and play.
- Create a virtual environment (replace 'project' with your prefreed pathname)
python3 -m venv project_env
- Activate your virtual environment as follows
source project_env/bin/activate
- Clone repository into the virtual environment created With git already installed, run the command below in your terminal to get the code into your local environment
git clone https://github.com/agbleze/AutoML_application.git
Make sure you are in the directory of the cloned repo, now you can install the package
pip install .