Scribd
Upload
9 views

Project

The document presents a tourism demand forecasting model for Tirumala Tirupati Devastanam (TTD) using machine learning, specifically a Light Gradient Boosting Machine Regressor. It incorporates various factors such as weather conditions, Google Trends, and Twitter sentiment to improve accuracy in predicting tourist arrivals. The study highlights the importance of effective forecasting in enhancing decision-making and resource allocation within the tourism industry.

Uploaded by

munigetigirish123
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
9 views

Project

The document presents a tourism demand forecasting model for Tirumala Tirupati Devastanam (TTD) using machine learning, specifically a Light Gradient Boosting Machine Regressor. It incorporates various factors such as weather conditions, Google Trends, and Twitter sentiment to improve accuracy in predicting tourist arrivals. The study highlights the importance of effective forecasting in enhancing decision-making and resource allocation within the tourism industry.

Uploaded by

munigetigirish123
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 23

Tourist Demand Forecasting

using Machine Learning


UNDER THE GUIDANCE OF
Mr. Kumar Anurupam Sir
ABSTRACT:
Our tourism demand forecasting model to forecast arrivals for the
tourist place Tirumala Tirupati Devastanam(TTD). We have considered
attributes Day Speciality(weekend or non-weekend ),Weather
condition(Based on temperature), Google Trends(frequency of ttd
related keywords searched on goolgle),Twitter trend(whether opinion of
twitter is +ve, -ve or neutral). Our model uses structured variables to
build a tourism demand forecasting model based on Light Gradient
Boosting Machine Regressor.
INTRODUCTION:

Motivation for the work:


Tourism industry contributes significantly to economic growth worldwide.

Forecasting tourist volume is crucial for predicting future economic


development.

Tourism demand forecasting aids in planning, policy making, and resource


allocation.

Predicting tourist arrivals helps destinations and businesses make efficient


decisions
CONT...
Real-word Application:

The project's real-world applications span destination


management, marketing, revenue management, policy making, and safety.

Forecasting tourist arrivals enhances decision-making processes and


contributes to the sustainable growth of the tourism industry.

The outcomes of the project can be implemented by various stakeholders


to optimize their operations and provide better experiences for tourists
RELATED WORK/EXISTING WORK

Previous research in tourism demand forecasting has investigated different


methodologies and variables. Some studies have focused on utilizing Google Trends data
and comparing accuracy across countries and cities. Others have explored the use of web
search index and image data for forecasting tourism demand. In this project, the objective
was to forecast tourist arrivals at the Tirumala Venkateswara Temple (TTD) using search
engine data, Twitter trends, historical data, and weather conditions. The proposed
approach involved employing a Light Gradient Boosting Machine Regressor (LGBMR)
model, incorporating these variables to enhance the accuracy of tourism demand
forecasting.
CONT…

LIMITATIONS:
Here we have concern in the availability of the dataset.Due lo lack of good dataset
many proposed models didn`t performed upto the level.

Spreading the forecasting for the multiple destinations is not possible.


PROPOSED METHOD:
This study used regression models to forecast the number of pilgrim arrivals at TTD,
focusing on correlated keywords like Tirupati, Tirumala, VIP darshan, and tirumala darshan.
Various machine learning models have been explored for tourism demand forecasting, including
using Google Trends and comparing accuracy between countries and cities. The proposed
approach involved using data from Google Trends, historical data, weather data, and Twitter
trends, with a Light Gradient Boosting Machine Regressor (LGBMR) model yielding accurate
forecasts. The process involved data collection, preprocessing, feature selection, training,
validation, and output prediction
LGBMR MODEL:
CONT…

LightGBM uses a leaf-wise tree growth strategy, which differs from


the level-wise strategy employed by many other gradient boosting
implementations. In leaf-wise growth, the algorithm selects the leaf node
with the maximum delta loss (improvement in the loss function) as the next
node to grow. This approach leads to faster convergence and better
accuracy.
CONT…
LightGBM optimizes the gradient descent algorithm by using a

technique called Gradient-Based One-Side Sampling (GOSS) and

Exclusive Feature Bundling (EFB). GOSS selects a small percentage of

large-gradient data instances and the full set of small-gradient instances,

resulting in a more efficient gradient approximation. EFB combines

exclusive feature values to reduce memory usage and speed up training.


EXPERIMENTAL SETUP

Technologies used:
● Selenium | LGBM algorithm
● Matplotlib | Pandas
● Scikit-Learn | Numpy
● Sentiment Analyzer | Jupyter Notebook
● Seaborn | plotly | Matplotlib
CONT…

DATASET:

The dataset used in this project includes historical tourism demand


records, Google Trends data, tweets related to the destination, weather
conditions, and weekday information. The dataset is carefully curated,
structured, and updated to capture various factors that influence tourism
demand. Data preprocessing techniques were applied to handle missing values
and outliers. The dataset serves as a foundation for training and evaluating the
tourism demand prediction model, providing valuable insights for accurate
forecasting
CONT…

SYSTEM HARDWARE:
● Modera Operating System
● 4 GB RAM
● 10 GB Free Disk Space
● X86 64-bit CPU(Intel/AMD Architecture)
RESULTS:

We evaluated four machine learning algorithms - LightGBM, Decision Tree,


Regression, and XGBoost - to predict tourist demand for TTD. LightGBM
performed the best, with the lowest MAE and RMSE values and the highest R^2
value. The results highlight the potential of our model in forecasting tourist demand
for TTD and assisting in resource allocation and crowd management. Further
refinement may be needed for improved accuracy
CONT…
CONTD…

RESULTS COMPARISION:
CONCLUSION:

Our tourist demand prediction project for Thirumala Thirupathi Devasthanam


(TTD) utilized the Selenium framework and effective preprocessing techniques.
LightGBM showed superior accuracy in predicting tourist demand. This
highlights the importance of machine learning in optimizing resource allocation
and enhancing visitor experiences at TTD. Refinement and collaborations will
advance tourist demand prediction, contributing to TTD's growth and success.
FUTURE SCOPE
Geometric Location Analysis: Incorporate geometric location as a factor in the prediction model
to analyze the impact of transportation connectivity, city development, and proximity to other
tourist destinations on tourist demand.

Integration of Additional Social Networking Data: Improve the model's performance by


incorporating reliable social networking data sources such as Instagram or Facebook to capture
user-generated content, sentiment analysis, and social engagement metrics for predicting tourist
demand.
COND…
Prediction of Multiple Destinations: We want to improve our model to predict the tourist arrivals
for the multiple destinations.Due to lack of dataset avaliability we are restricting it to TTD only.

Collaboration with Tourism Stakeholders: Collaborate with tourism authorities, travel agencies,
and other stakeholders to gather insights, incorporate feedback, and ensure the model aligns with
industry needs for better decision-making and enhanced visitor experiences
REFERNECES:
https://drive.google.com/drive/u/0/folders/1D79qkvtBdHIS2uV6PeixbfX6JA5c--hf

https://drive.google.com/drive/u/0/folders/1Fe411WLN8xl6vzlC_1Ftznej6_PSjy0k

https://drive.google.com/drive/u/0/folders/1wylQ4EkjJov9jJpi4ZUuWgSf_0YVF5KS

https://drive.google.com/drive/u/0/folders/19gfLumZbdaDNKg3gHXhui6crtOv6GG
TS

https://drive.google.com/drive/u/0/folders/1IeQLpFyh0XlICgdePwQzV5qG4VCf4m
Zv
THANK YOU EVERYONE

You might also like