Scribd
Upload
255 views8 pages

Forest Fires Data Set Analysis Using Machine Learning: Name: 1.pawan Jakke (111815018) 2.utkarsh Dubey (111815047)

The document discusses analyzing a forest fires dataset using machine learning algorithms. It describes past research that used various machine learning methods like SVM to predict burned forest area. The objectives are to perform data analysis on the forest fires dataset and build a regression model to predict burned area. So far, the authors have learned basic machine learning in Python and will analyze the dataset to train and test models using different algorithms to find the most accurate one. Their next steps are to create a model with the best performing algorithm and analyze the full dataset to achieve high prediction accuracy.

Uploaded by

Aman Dubey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
255 views8 pages

Forest Fires Data Set Analysis Using Machine Learning: Name: 1.pawan Jakke (111815018) 2.utkarsh Dubey (111815047)

The document discusses analyzing a forest fires dataset using machine learning algorithms. It describes past research that used various machine learning methods like SVM to predict burned forest area. The objectives are to perform data analysis on the forest fires dataset and build a regression model to predict burned area. So far, the authors have learned basic machine learning in Python and will analyze the dataset to train and test models using different algorithms to find the most accurate one. Their next steps are to create a model with the best performing algorithm and analyze the full dataset to achieve high prediction accuracy.

Uploaded by

Aman Dubey
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Forest Fires Data Set Analysis

using Machine Learning

Name : 1.Pawan Jakke


(111815018)

2.Utkarsh Dubey Supervisor: Dr.Sanjeev Sharma.


(111815047)
Introduction
In [Cortez and Morais, 2007], the output 'area' was first transformed with a ln(x+1) function.

Then, several Data Mining methods were applied. After fitting the models, the outputs were

post-processed with the inverse of the ln(x+1) transform. Four different input setups were

used. The experiments were conducted using a 10-fold (cross-validation) x 30 runs. Two

regression metrics were measured: MAD and RMSE. A Gaussian support vector machine (SVM) fed

with only 4 direct weather conditions (temp, RH, wind and rain) obtained the best MAD value:

12.71 +- 0.01 (mean and confidence interval within 95% using a t-student distribution). The

best RMSE was attained by the naive mean predictor. An analysis to the regression error curve

(REC) shows that the SVM model predicts more examples within a lower admitted error. In effect,

the SVM model predicts better small fires, which are the majority.
Literature Survey
Title Methodology Published year
Development of a Surveillance System for Forest Fire Detection The purpose of the project is to detect a forest fire in its earliest
and Monitoring using Drones stage possible by implementing a surveillance system which
2019
primarily consists of drones. The nearest authority will be
notified shortly about the location of the fire.
Md. Saif Allauddin ; G. Sai Kiran ; GSS. Raj Kiran ; G Srinivas ;
G Uma Ratna Mouli ; P Vishnu Prasad

Prediction of Forest Fire using Neural Network based on The method used in this study is a neural network with Extreme
Extreme Learning Machines (ELM) Learning Machines (ELM) training model. To improve the 2019
performance of the ELM method, in this study several tests will
be carried out so that the resulting predictions are the best.
Mochammad Anshori ; Farhanna Mar'i ; Mukhammad Wildan
Alauddin ; Wayan Firdaus Mahmudy

An Integrated Approach to the Regression Problem in Forest The developed approach was tested on the data of
Fires Detection Montensinho park. In this paper developed approach is 2020
compared with classic regression algorithms and its advantages
are shown. The significantly better quality of the proposed
I. S. Sineva ; M. D. Molovtsev solution is demonstrated.

Classification Algorithms Analysis in the Forest Fire Detection In this paper the analysis of ML algorithms and the possibility of
Problem their application to forest fires data are done. 2019

Mikhail D. Molovtsev ; Irina S. Sineva


Objectives

1. Data Analysis is a process of inspecting, cleansing, transforming


and modeling data with the goal of discovering useful information,
informing conclusions and supporting decision-making.

2. Forest fires cause a significant environmental damage while


threatening human lives. The database included spatial, temporal,
components from the Canadian Fire Weather Index (FWI) and four
weather conditions. This problem was modeled as a regression
task, where the aim was the prediction of the burned area.
Methodology
Unlike other fire detection methods, this framework is not designed to detect fire through fire
features. It has been shown that focusing on fire features often reduces the stability and
reliability of a fire detection system as fire features are easily confused with natural
phenomena [21]. Therefore, instead of looking for fire features in images based on colour
and motion, this study aims to train a program with sets of fire and non-fire images from the
forest and learn the program by introducing a feature vector based on similarities between
features of the data. The method includes:

• Framework Construction and Implementation in Matlab

• Data Collection

• Training the Model

• Testing the Model

• Performance Analysis of the Model


Work Done So far:

We have learnt basics of Machine Learning with Python. Now we are going
to take dataset and we will do analysis to that dataset.

We pick an Algorithm and find accuracy.

Train the model

Test the model

Compare Test with predictions finally


Target for next week:
● We are building a model. And we are taking any dataset, Splitting Data
set into two parts as Training and Testing. Then we assign as 80% for
training and 20% for Testing. And predicting the accuracy of the model
with various Algorithms. Now if we give any input the model must predict
the Input given and gives the suitable output as the accuracy goes on
increases.

● We will create a model with an algorithm which gives better accuracy


comparing to other algorithms. We will analyze full dataset and get good
accuracy with our model.
References
1. M. S. Allauddin, G. S. Kiran, G. R. Kiran, G. Srinivas, G. U. R. Mouli and P. V. Prasad,
"Development of a Surveillance System for Forest Fire Detection and Monitoring using Drones,"
IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium, Yokohama,
Japan, 2019, pp. 9361-9363, doi: 10.1109/IGARSS.2019.8900436.
2. M. Anshori, F. Mar'i, M. W. Alauddin and W. F. Mahmudy, "Prediction of Forest Fire using Neural
Network based on Extreme Learning Machines (ELM)," 2019 International Conference on
Sustainable Information Engineering and Technology (SIET), Lombok, Indonesia, 2019, pp.
301-305, doi: 10.1109/SIET48054.2019.8986106.
3. I. S. Sineva and M. D. Molovtsev, "An Integrated Approach to the Regression Problem in Forest
Fires Detection," 2020 Systems of Signals Generating and Processing in the Field of on Board
Communications, Moscow, Russia, 2020, pp. 1-4, doi: 10.1109/IEEECONF48371.2020.9078645.
4. M. D. Molovtsev and I. S. Sineva, "Classification Algorithms Analysis in the Forest Fire Detection
Problem," 2019 International Conference "Quality Management, Transport and Information
Security, Information Technologies" (IT&QM&IS), Sochi, Russia, 2019, pp. 548-553, doi:
10.1109/ITQMIS.2019.8928398.

You might also like