Weather Analysis and Prediction

A
Mini Project Report
Submitted
by

Ms. Shivani Ravindra Divekar 2041016

Ms. Sanika Sunil Patale 2041057

In partial fulfillment for the requirement of Laboratory Practice-II

of

BaĚelor of Computer Engineering

Under the guidance of

Digambar Padulkar
(Assistant Professor)

Department of Computer Engineering

Vidya Pratishthan’s Kamalnayan Bajaj Institute of
Engineering and Technology
Bhigwan Road, Vidyanagari
Baramati-413133
2020-2021
 Vidya Pratishthan’s
Kamalnayan Bajaj Institute of Engineering and Technology, Baramati
Department of Computer Engineering

Certificate
This is to certify that following students
Ms. Shivani Ravindra Divekar 2041016
Ms. Sanika Sunil Patale 2041057

have successfully completed their project work on

Weather Analysis and Prediction
during the academic year 2020-2021 in the partial fulfillment towards
the completion of Laboratory Practice-II in Computer Engineering.

Project Guide HoD Deptt. of Comp. Engg.

(Digambar Padulkar) (Dr. Santaji Shinde)

Principal
(Dr. R. S. Bichkar)

Internal Examiner External Examiner

 Acknowledgments

We are extreamly grateful to Dr. R.S. Bichkar,Principal,VPKBIET and Dr. S.K.Shinde

HOD.Head of Department,Department of computer Engineering for providing all the re-
quired resourses for the successful completion of our mini project. We express our heart-
felt gratitude to our project guide Prof.Digamber Padulkar.Assistant Professor,Computer
Engineering,for his valuable suggestions and guidance in the project report. Last but not
least we are very much thankful to our parents who guided us in every step which we
took.
Ms. Shivani Ravindra Divekar
Ms. Sanika Sunil Patale

i
 Abstract

Weather Prediction and Analysis ia an important application in meterology and

has been one of the most scientifically ana technology challenging problem around the
world.In this project,we investigatet the use of data mining techniques in forecasting max-
imum temperature,rainfall,evaporation and wind speed.Weather prediction approches are
challenged by complexed weather phenomenon with limited observations and past data.

ii
 Contents

Acknowledgments i

Abstract ii

1 Introduction 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Brief Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Dataset Description 2
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1.1 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1.2 Project Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

3 Data Preprocessing 3

4 Classification 5
4.1 KNN Classifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.2 Logistic Regression Classifier . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.3 Decision Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

5 Confusion Matrix 7
5.1 Analyse Confusion Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . 7

6 Result Analysis and Conclusion 9

iii
 1

Introduction

1.1 Overview

This section of project provide overall description of the project,including brief description
and problem defination.Second section contains literature survey. Third section contain
introduction ,purpose and design and implementation constraints.fourth section contain
classification and fifth section contain Confusion Matrix including Analysis of Confusion
Matrix.

1.2 Brief Description

Weather forecast are made by collecting as much data possible about the current state of
atmosphere and using understanding of atmospheric process to determine how atmosphere
evolves in the future.In this project we have taken dataset as a input which consist
of different attribute like temperature, humadity,sunshine,rainfall etc.In this project we
predict weather for prediction we appiled different classification technique and compare
the accuracy for the best result.

1.3 Problem Definition

Cosidering a weather dataset we predict which changes happen in atmosphere at partic-

ular time.For this purpose we use here different classification techniques.

1
 2

Dataset Description

2.1 Introduction

2.1.1 Purpose

Weather is one of the most effective environmentel constarints in every phase of our
lives.We are subject to adjusting ourselves with respect to weather condition from our
dressing habits to stategic organizational planning activities.The main purposeb of this
project is predict accurate enviornment condition.

2.1.2 Project Scope

Weather Aanalysis and Prediction is the application of science and technology to predict
the state of location. Design and Implementation Conthe atmosstraints

2
 3

Data Preprocessing

Data Cleaning -
Data cleaning is the important process of Data pre-processing.It
is the process of detecting and correcting corrupted or inaccu-
rate records from a record set.For example filling missing value
,smooth noisy data, identify or remove outliers etc.

3
 CHAPTER 3. DATA PREPROCESSING

Label Encoding -
It refers to converting the label into numeric from so as to convert
it into machine readable form.

Weather Analysis and Prediction 4 VPKBIET, Baramati

 4

Classification

Classification is the process of predicting the class of given data

points.It belong to the supervised learning category where the
targets provided with input data.Classes are sometimes called as
label or categories.

4.1 KNN Classifier

K-nearest algorithm which is a non-parametric method used for

classification and regression.Output of this classification is a class
membership.Our goal of this classifier is to find groups in the data
and that groups are represented by K variable. Accuracy of Knn
classifier on the training dataset is 0.89.

5
 CHAPTER 4. CLASSIFICATION

4.2 Logistic Regression Classifier

Logistic regression used for analysis a data set.It is statistical

method in which one or more independent variables that deter-
mine outcome.

4.3 Decision Tree

Decision tree learning is method commonly used in data mining.

The goal is to create a model that predicts the value of a target
variable based on several input variable. It is atree in which each
internal node is labeled with an input. this is easy to understand
and interpreted.

Weather Analysis and Prediction 6 VPKBIET, Baramati

 5

Confusion Matrix

5.1 Analyse Confusion Matrix

To describe performance of the classification confusion matrix

is used.It is also known as error matrix. It contain table having
dimension actual and predicted.Each row of this matrix describes
the instance in a predicted class while each coloumn represents
the cases in actual class.
Confusion matrix for Knnc

7
 CHAPTER 5. CONFUSION MATRIX

Confusion matrix for Logistic Regression

Confusion matrix for Decision Tree

Weather Analysis and Prediction 8 VPKBIET, Baramati

 6

Result Analysis and Conclusion

To predict whether tomorrow will rain, today’s humidity and

rainfall(MM) is very important, higher humidity or higher rain-
fall today will lead to higher probability rain tomorrow. Selecting
the most relevant features helps to make the modeling process
more efficient, and make it easier to understand the prediction
results. But compared to dimension reduction, feature selection
will lose some information due to dropping some features com-
pletely.