Machine Learning Predicts Floods and Landslides [2024] | AI Project

Summary

The video script introduces 'Flood Prediction Using Machine Learning,' an innovative project aimed at saving lives in flood-prone areas like India. It leverages Convolutional Neural Networks (CNN) to analyze vast datasets including weather reports, river levels, and satellite images for accurate flood and landslide prediction. The project boasts a 97% accuracy rate and provides real-time monitoring and predictions through a user-friendly website, potentially aiding in disaster preparedness and response.

Takeaways

• 🌊 The project aims to predict floods using machine learning, which is crucial for saving lives, especially in disaster-prone areas like India.
• 🔮 The system uses convolutional neural networks (CNN) for analyzing large datasets including weather reports, river levels, satellite images, and historical data to predict flood events with high accuracy.
• 📈 The CNN model boasts an impressive 97% accuracy rate in real-time predictions, which is a significant improvement over the 74% accuracy of the previous model using LSTM.
• 📚 The project considers various data inputs such as historical rainfall, current rainfall, dam levels, and satellite imagery for training the model.
• 🔍 The model identifies correlations and patterns from the data to predict the likelihood of floods and landslides based on current conditions.
• 🏭 The project also addresses the issue of data scarcity by using APIs to collect comprehensive data from different regions.
• 🌐 The system integrates with a website built using Flask, allowing users to input a location and receive flood predictions and alerts.
• 📊 The website provides dynamic graphs and heat maps for visualizing flood-prone areas, flood intensity, rainfall predictions, and potential financial losses due to floods.
• 🛠️ The project's architecture includes modules for data collection, pre-processing, model training, and prediction, with Flask used for the user interface.
• 💻 The project can be run on minimum I3 processors and requires Python, making it accessible for implementation on various platforms.
• 🔗 The project's success is demonstrated through its high accuracy, real-time monitoring capabilities, and the ability to predict not only floods but also landslides.

Q & A

What is the primary goal of the 'Flood Prediction using Machine Learning' project?

The primary goal of the project is to save millions of lives by predicting floods and landslides using machine learning techniques, providing hope and safety to communities at risk.

How does the project address the issue of data analysis for flood prediction?

The project uses machine learning techniques to analyze large amounts of data in real time, including weather reports, river levels, satellite images, and historical data, which would be challenging to process manually.

What machine learning algorithm is central to this project?

The project primarily uses Convolutional Neural Networks (CNN) as the trained algorithm for analyzing the data and making predictions.

What kind of data is considered as input for training the CNN model in this project?

The input data for training the CNN model includes historical rainfall data, current rainfall data, dam levels, and satellite images, which are all part of the dataset considered for analysis.

How does the CNN model determine the likelihood of a flood or landslide?

The CNN model identifies relevant features from the dataset, considers correlations and factors that contribute to flood or landslide occurrences, and compares current levels with the trained dataset to predict the likelihood of such events.

What is the reported accuracy of the CNN model used in the project?

The CNN model used in the project has achieved an accuracy of 97%, which is a significant improvement over the existing systems.

How does the project differentiate from the existing system that uses LSTM?

The existing system using LSTM achieved only 74% accuracy and suffered from drawbacks such as low accuracy and high computational complexity. The proposed system in the project overcomes these issues with a more accurate CNN model and additional data considerations.

What are the main challenges addressed by the proposed system in the project?

The main challenges addressed include data scarcity, computational complexity, and model interoperability, which were limitations in the existing system that the project aims to overcome with its approach.

What are the additional factors considered in the proposed system for improved flood and landslide prediction?

The proposed system considers factors such as satellite imagery, rainfall patterns, topographic patterns, and soil moisture levels, in addition to historical weather patterns and dam levels.

How does the project provide real-time monitoring and prediction?

The project integrates real-time data collection through APIs, pre-processes the data, and uses it for feature engineering. The CNN model is then trained on these features, and the Flask framework is used to create a website for real-time monitoring and prediction.

What is the significance of the project's ability to predict not just floods but also landslides?

The ability to predict both floods and landslides provides a more comprehensive approach to disaster management, allowing for better preparedness and mitigation strategies in areas susceptible to these natural disasters.

How does the project's website facilitate user interaction with the flood and landslide prediction model?

The website allows users to input the name of any city in the world, fetches real-time data for that location, and then uses the machine learning model to predict the risk of flood or landslide, providing immediate and location-specific insights.

Outlines

00:00

🌊 Innovative Flood Prediction with Machine Learning

This paragraph introduces a project on flood prediction using machine learning, emphasizing its potential to save lives in India. The project aims to analyze large datasets including weather reports, river levels, and satellite images in real-time using machine learning techniques. It specifically uses Convolutional Neural Networks (CNN) for training with historical and current data to predict flood occurrences and their severity. The CNN model boasts a high accuracy rate of 97%, enabling precise predictions of when and where floods will happen, and even assessing potential landslides. The project's significance is underscored by the loss of over 5,000 lives due to floods, highlighting the urgency of accurate flood prediction.

05:00

📈 Enhancing Flood and Landslide Prediction with Advanced Models

The second paragraph delves into the project's approach to predicting floods and landslides, using a combination of historical weather patterns, dam levels, land usage, and socio-economic indicators. The project's goal is to efficiently predict the extent of natural disasters and their potential for wealth loss. It contrasts the proposed machine learning approach with the existing system, which relies on hydrological models and suffers from data scarcity and computational complexity. The proposed system overcomes these issues by integrating various data sources and using CNN for higher accuracy predictions. The system is designed to be continuously trainable and upgradeable, with real-time monitoring and predictive capabilities.

10:03

🛠️ Project Architecture and Implementation Details

This paragraph outlines the technical aspects of the flood prediction project, detailing the architecture and the sequence of models involved, including data collection, pre-processing, model selection, and prediction. It also discusses the project's diagrams such as class, activity, and use case diagrams. The accuracy of the proposed model is highlighted, showing a significant improvement over the existing model, with a 97% prediction rate and an F1 score of 99. The hardware requirements for running the project are mentioned, and the paragraph concludes with instructions on how to run the project using Python and Flask, resulting in a web application that can predict flood risks based on real-time data.

15:04

🌐 Real-time Data Integration and Dynamic Prediction Visualization

The fourth paragraph focuses on the real-time data integration and dynamic visualization of the flood prediction project. It describes the use of heatmaps and satellite images to assess potential losses in terms of USD and the impact of floods in metropolitan areas. The project's ability to generate dynamic graphs based on live data is emphasized, allowing users to view flood intensity, rainfall predictions, and other relevant metrics. The paragraph also explains how the project can predict flood risks for any city worldwide, providing a comprehensive and interactive user experience through its web application.

20:06

📊 Demonstrating Flood Prediction Capabilities with Global Cities

This paragraph demonstrates the global applicability of the flood prediction project by showing how it can provide real-time data and predictions for various cities, including Chennai, Delhi, Mumbai, Hyderabad, and even international cities like London and New York. It illustrates the process of obtaining latitude and longitude data, fetching real-time weather information, and using this data to perform machine learning operations that predict flood risks. The project's ability to provide detailed weather insights and flood safety categories for different cities is highlighted, showcasing its versatility and accuracy.

25:07

🏆 Achieving High Accuracy with Machine Learning in Flood Prediction

The final paragraph wraps up the project by emphasizing its high accuracy, which is around 97%, and the importance of the features considered for training the machine learning model, such as drainage value, dam quality, river management, political factors, and population score. It also provides an overview of the project's overall architecture and invites interested parties to visit the I Expert website for more projects on machine learning and blockchain, indicating the broader scope of available solutions.

Keywords

💡Flood Prediction

Flood prediction refers to the process of anticipating and estimating the likelihood and impact of flood events. In the context of the video, it is a crucial theme as it discusses a machine learning project aimed at saving lives by predicting floods. The script mentions the importance of this technology in India, where floods have resulted in significant loss of life, highlighting the project's goal to build hope for communities by leveraging machine learning techniques.

💡Machine Learning

Machine learning is a subset of artificial intelligence that enables computers to learn from and make decisions based on data. The video's theme revolves around using machine learning for flood prediction. The script describes how machine learning algorithms can analyze vast amounts of data, such as weather reports and river levels, to predict flood events with high accuracy, thus serving as a life-saving technology.

💡Convolutional Neural Network (CNN)

A Convolutional Neural Network is a type of deep learning algorithm widely used in image recognition and processing tasks. In the video script, CNN is highlighted as the trained algorithm used in the flood prediction project. It extracts relevant features from datasets, such as historical and current rainfall data, dam levels, and satellite images, to predict flood occurrences with a high degree of accuracy.

💡Data Analysis

Data analysis involves inspecting, cleaning, transforming, and modeling data with the goal of discovering useful information. The script emphasizes the need for analyzing various data types, including weather reports, river levels, and satellite images, to predict floods. The project leverages the power of machine learning to handle and analyze these large datasets in real-time.

💡Accuracy

In the context of machine learning models, accuracy refers to the proportion of correct predictions made by the model. The video script boasts the CNN model's high accuracy of 97% in predicting floods, which is a significant achievement as it allows for precise flood forecasting, potentially saving millions of lives.

💡Historical Data

Historical data refers to information from past events or periods, which can be used to inform predictions about the future. The script mentions using historical rainfall data as part of the dataset for training the CNN model. This data helps the model understand patterns and correlations that may lead to flood events.

💡Real-time Data

Real-time data is information that is processed and analyzed as it is collected, without any delay. The video script discusses the importance of real-time data in the flood prediction project, as it allows for current conditions to be assessed and compared against the model's trained dataset to make immediate flood predictions.

💡Landslide Prediction

Landslide prediction is the process of forecasting the occurrence and impact of landslides, similar to flood prediction. The script mentions that the machine learning project not only predicts floods but also landslides, expanding the scope of the technology to include another natural disaster with significant potential for damage and loss of life.

💡Hydrological Data

Hydrological data pertains to information about water in its various forms and cycles, including data on water levels in dams and rivers. The script describes how hydrological data is one of the key datasets used in the flood prediction model, providing insights into water levels that could indicate an impending flood.

💡API

An API, or Application Programming Interface, is a set of rules and protocols that allows different software applications to communicate with each other. The video script refers to using an API to collect real-time data for the flood prediction project, emphasizing the importance of integrating external data sources to enhance the model's predictive capabilities.

Highlights

Introduction of a flood prediction project using machine learning with potential to save millions of lives in India.

Flood prediction is not only a technological advancement but also a beacon of hope for communities and the future.

The project emphasizes the importance of flood prediction after the loss of 5,12 people due to floods, highlighting its urgency.

Data analysis for flood prediction involves weather reports, river levels, satellite images, and historical data.

Machine learning techniques, specifically convolutional neural networks (CNN), are used to analyze large datasets efficiently.

The CNN algorithm is trained with historical and current data, including rainfall and dam levels, to predict flood patterns.

The project boasts a high accuracy rate of 97% in real-time flood prediction, a significant advancement in the field.

The project's base paper discusses the use of LSTM for flood prediction but achieved lower accuracy at 74%.

The proposed system aims to overcome the existing system's drawbacks, such as data scarcity and computational complexity.

The project integrates satellite imagery, rainfall patterns, topographic patterns, and soil moisture levels for comprehensive flood and landslide prediction.

The system is designed to be continuously trained and upgraded to improve accuracy over time.

Real-time monitoring and predictive modeling are included in the project to provide immediate and accurate flood alerts.

The project's architecture includes data collection, pre-processing, model selection, and predicting models for a streamlined process.

The project provides a user interface through a website, allowing users to input area names and receive flood alerts.

The system uses Flask to build a website for user interaction and real-time data fetching.

Heat maps are used to estimate potential financial losses due to floods, providing insight into disaster management.

The project's code is open-source, allowing for community contributions and further development.

The project concludes with a demonstration of its capabilities, showcasing its potential impact on saving lives and resources.

Transcripts

hello everyone welcome to I expert today

we are going to see one of the

interesting project which is named as

flood prediction using machine learning

this is the very Innovative project

which could save millions of lives in

India Flint prediction not only a

technology blood prediction belongs to

people it's more about Hope communities

and future that's what we are going to

we are not only building a technology

using machine we are building hope for

the peoples so with the help of copes we

can save millions of lives in future so

that this project will be more important

for the current Society stand last are

we lost 5 12 people lives due to the

flood so this so you can see the

importance of this project okay FL

prediction is not a easy task you should

analyze various amount of datas for

example weather report River levels

satellite images historical plate

these are the datas you should consider

but these datas are huge in real time

but than to the power of Miss with the

help of this machine learning Technique

we can analyze these datas very easily

how we have to analyze these datas so

this project works on machine learning

in this project we are using

conversional neural network as a trained

algorithm you have to give various data

to train for example you have to give

historical data as a historical rainfall

data current rainfall data then Dam

levels as well as satellite images these

are the datas you have to consider to

give as a input while training these

datas will called as a data set what CNN

will do means it will take relative

features from the data set which you are

giving it will take while the FL

happening in the existing system it will

take what type of features what type of

serious correlation will happen for

example if the dam level was in this

level if the rain will happen in this

level Maybe the flood will happen this

correlation factor that will consider

the CNN will consider this correlation

Factor this relation factor from the DAT

to then what it will do means it will

Max the correlation factor with the

current real now what's the dam level

now what is the satellite imaginary

value now what is the rainfall value it

will compare this current level values

with the previous data set value then it

will find the relative propag

how much propagation of flood will

happen how much propagation of landslide

will happen that's what this D will do

the main accuracy of main advantage of

CNN here was it having 97% accuracy

which is huge in real time okay you can

predict the flood exactly when the flood

will happen where the flood will happen

how much amount of flood will happen

this thing you will predict very easily

the landslide also you can consider from

the blood level so this is the project

please connect with this video we will

explain complete demo of the project

with presentation with coding and

everything thank

you uh let me explain the project base

paper as well as presentation this is

our project base paper this is it based

base paper okay the title as spal

temporal flenders are mapping using

integration of telemetry data and

prediction model this is the project

base paper we are considered for this

project in the project base paper they

have taken lstm algorithm they have

taken lstm long shortterm memory this is

the algorithm they have used to train as

well as testing this project the main

drawback of this project was they

achieved very low accuracy as you can

see they achieved

74% accuracy the main drawback was this

one only 74% achieve accuracy only they

achieved for this project the

architecture of the existing system

shown

here they have

considered water level as well as hourly

water level as well as daily water level

these datas are trained under temporal

prediction model that is called lsdm

based on the gque they have estimated

the FL reject this is completely

existing system very low in accuracy to

overcome this drawback only we are going

for our propos system so this is our

propos system BPT as you can see the

title was flood and Landslide prediction

using machine learning we have created

flood as well as Landslide prediction

what the model we have created for this

project the abstract we have mentioned

that natural disaster such as Landslide

as well as blood are posed to

significant challenges in human regions

okay this stud present uh Innovative

machine learning approach to predict

this Landslide as well as flood

insecurity in suspectable area this is

the obstruct part of project in this

project we are using historical weather

patterns Dam levels land usage what is

the weather data socio Eon socio

economic indicators to predict the how

much loss will happen Okay how much

wealth loss will happen we are using

this five input in our data set to trade

okay in introduction we are given that

how frequently how severely this natural

disaster uh raised to significant

concert okay so we have to make sure

proposed algorithm able to predict this

flood as well as lands slate very

efficiently that is the major Moto of

this project okay why because India

contains uh India having more number of

human lives as well as India having more

number of forest when compared to other

regions so the flood as well as landfall

prediction last Light prediction is very

important for India okay the role of

machine learning here was the predictive

approach in this project we are using

predictive approach no for prediction

means you have to apply huge accuracy

High higher number of accuracy for that

purpose only we are using machine

learning here okay it don't use any data

mining and all it uses pure machine

learning approach the main focus of this

project was two things flood prediction

as well as last Light prediction these

are the main two focus of this project

the existing system existing system uses

three models hydrological models

hydrological models means the base paper

say water level they are solely used

water level to predict the flood but we

are using some Advanced consideration

okay the hydrological model remote

sensing some person using satellite maps

to predict the water flow at all okay

and rainfall prediction these are the

existing methodology used in our project

but the main drag B was Data scarcity

data scarcity means you can't get exact

data for all the region in existing

system but we are using proposed API

here the proposed API capable capable of

getting all the data for all the regions

okay we are using AP here let me discuss

everything on the project demo section

computational complexity the lstm pr to

complexity okay it can't handle all the

datas the main drawback of the LST was

computational complexity then model

interoperability model interoperability

means the Deep learning things was

difficult to interpret making challenge

to understand how the prediction of uh

predictions are made okay you can't

understand the data clearly that is the

main drawback of the exiting system

these drawbacks are overcome by our

propose system in our propose system we

are using we are going to predict flood

as well as last SL both the things we

have to predict okay this system

integrate satellite imaginary rainfall

pattern then topographic patterns as

well as soil moisture level R flow

levels these factors are considered to

to predict the flood as well as landfall

Direction so that this accuracy of the

project was huge when compared to the

existing system the main algorithm used

in this project was conversion neural

network this only achieved the 97% of

accuracy okay as well as uh the proposed

conversion neural network model can be

continuously trained as well as upgraded

okay over the time to improve the

accuracy so this system also include the

realtime monitoring as well as propos to

prediction model what the things

included in this project this is the

overall architecture of this project in

the architecture as I mentioned earlier

I'm using historical data sensor data

means what's the current temperature

what's the current rainfall rate as well

as current water level of the particular

dams these are the sensor data then

weather forecast weather forecast means

predictable forecast what the maybe a

rain will come tomorrow or not what's

the expected uh range of temperature for

tomorrow these are the data called as

weather forecast then hydrological data

means as I mentioned here what's the dam

level how much amount of water the L Dam

can survey these are the data called as

hydrological data these datas are

collected and these data are

pre-processed then we made final data

set from the data set we made feature

engineering selected features can be

pred from the data set that model that

features can be trained under the

particular CNN model okay that CNN will

train particular feature it store on the

particular model okay then we are using

flask model to validate this result we

are building particular website in the

website we can give you can give the

input particular area what type of what

are the area you have to consider you

can give the area name then what our

project will do means it will fetch the

realtime data from the for the

particular place this real time datas

are always compared to the already

pre-ra model based on the pre-ra model

probability it will give whether you are

having flood alert or not okay these

things are accessed by the user

interface as I mentioned earlier by

using flask we made uh beautiful website

for this project uh these are the four

models involved in the project data

collection model pre-processing model

then model selection then predicting

model these are the four stems involved

in this project okay models uh this is

the sequence diagrams of the project

class diagram activity diagram use case

diagram all the diagrams given for this

project if you're purchasing means you

can access all the things for free okay

uh you can see the result okay the

proposed model having accuracy of 97% as

I mentioned ear existing model having

accuracy of 71% the pr rate was PR rate

of propos system was 97 as well as

exitting system having press rate of 70

only the proposed system having uh F1

score of 99 exit having F score of 81 so

you can see the accuracy of our project

which is improved by huge margin

this is a hardware requirement of the

project you need minimum I3 process to

run this project okay in this project

can be run on Windows as well as Mac OS

we need python to run this project in

conclusion we have given that this is

the interesting project with the help of

CNN we have improved to improv the

prediction model of flood as well as

Landslide it poses significant

advancement when compared to the base

paper it we are accessing vast amount of

data set to make the accuracy as well as

the data set contains historical weather

pattern topography as well as oil

condition we are using machine learning

model that's called CNN to apply the

predictive foress reference for this

project okay uh please connect to this

video for demon sessions let's move on

to the project demo so this is the pro

uh this is the project code folder you

can find the complete coding here this

is already pre-trained model this is our

python code which you are using this is

the training code which are used to

predict the project then this is the

template folder the template folder we

have used frontend HTML pages okay to

run this project you have to copy the

project code location then open Ana

Navigator just click python

terminal once you open the python

terminal means use CD space paste the

project location then enter to run the

main project you have to type pythons

space app.py this is the command used to

run the main project so once you type

the python space app.py means it start

run the

project so it's executed now so this is

your local host address just copy the

local host address paste it on the

browser so this is our project homepage

this website running from our terminal

okay this terminal on generated this

website with the help of these codes

okay so you can find the complete

project on the Local Host to start this

project this is our homepage we named

our project as a flood guard flood guard

powered by AA this is our project code

folder sorry this is our project

homepage we have created some homepage

about our project about us everything

you have mentioned then what are the

services we are providing means plot

heat Maps satellite images predict flood

these are the services included in this

project so then why this project

important means this project will help

on the first date rescue operation also

this project we can estimate to predict

how much food supply uh person will

demand or while flood happening so this

is our contactor form then some

frequently Asked question also we have

displayed this website completely

running with the help of python okay

this website completely running with the

help of python the first step which are

going was plots in the plots you can see

the complete flood areas okay flood

prediction what are the

places what are the places will be flood

prawn in India you can see the complete

results okay can see the complete

results in the red colored things

referred to flood occurring PE and flood

occurring places okay the number belongs

to latitude and longitude position this

website this graph automatically

generated with help of our prediction

you can zoom

also if you are zooming here means you

can get the clear idea for example

Mumbai PR to flood then also you can get

Chennai Chennai also PR to flood then Ki

then tanur these are the places PR

flood these are the places okay then the

second graph belongs to how much flood

will happen in the each and every place

okay how much flood will happen on the

each place this graph completely Dynamic

graph will be generated based on our

prediction process okay you can see the

flood intensity this is flood intensity

graph you can see this is these are the

places flood may happen in the huge data

okay this is the

fled intensity graph okay and this is

the flood intensity graph then the third

graph will

be how much rain will happen on the each

and every places how much rain will

happen on the each and every place you

can see the rain plot here okay these

are this also will be generated based on

our plot okay prediction plot then go

for the heat

Maps heat Maps directly related to our

satellite so this is the heat map heat

map belongs to how much loss will happen

in terms of US dollars so if fled

happened means how much loss will happen

in terms of you uh US Dollars this also

very much important why because if the

flood happening in Forest places mean

there is no huge loss for few months but

if the flood happening on the Metro City

means the loss ratio will be higher the

heat map will uh uh help us to find out

how much loss how much money loss will

happen in terms of flood as well as

rainfall okay you can see if the flood

happening on the New Delhi means this is

the New Delhi surrounding if the flood

happening on the New Delhi surrounding

means the loss ratio will be huge also

the flood will happening on the Mumbai

side means the loss will be huge why

because these areas are surrounded by

various number of peoples okay huge

population that's why it it will make

huge loss okay that's Al important also

in hit M you can get Chennai PR to less

loss when compared to Delhi why because

Chennai population count is different as

well as Chennai architecture count is

Chennai architecture style is different

okay as well as Chennai CV policy also

different so we compared to Chennai

Delhi will meet huge loss for the same

amount of uh rain okay then partiti

participation also refer to the PO

population ratio and loss ratio here

also you can find out Deni Mumbai then

this place West Bengal these places are

PR to flood as well as huge loss if some

rain will happening means

okay so this also key map then directly

I can go for the satellite these graphs

are completely generated by our

prediction process these graphs are

completely Dynamic okay it will change

automatically so for example if you're

going for D in July Monthan June month

means you can see the rain data here in

rain data these are the data rainfall

data from Deli if the same thing I'm

turning into July means you all know in

July three three or 4% died on uh Delhi

you can see the complete graph compared

to June you got huge rainfall in Delhi

surrounding also you can change the city

here for example I taken Mumbai in May I

have taken we can see see the result May

Mumbai got low rainfall so I can turn it

in July for Mumbai so compared to May in

July month Mumbai got huge rainfall this

is the cloud ratio from where clouds are

gathered near to Mumbai okay I can

change into Chennai also these are the

Metro City I can change so Chennai got

less amount of rainfall compared to

other Metro City this is the important

page in our project prediction page in

prediction page you can type any city in

the world you can type any city in the

world what this project will do means it

will fetch the lude and longitude

portion of the uh particular City it

will get the weather data from this AP

visual Crossing AP this project Real

Time Project it you can check this

project any date anything okay it will

work perfectly this is the AP we have

used for used to collect the data real

time data for example I'm using chenai

here means it will give the chenai

results okay this is the chenai result

okay chenai temperature uh rain days

what's the wind speed and all you can

get forecast for 15 days uh 40 days

everything you get yesterday

today so what our project will do means

in our project we to get complete graphs

from this website using this graph it

will perform the machine learning

operation it will perform the machine

learning operation for the current data

live data it will collect live data from

this website based on the live data it

will predict whether the rainfall or the

last may happen it will predict that I'm

just typing chenai here so before that

you can check the command box if I'm

typing chenai here means what are my

what my project will do means it will

get the latitude longitude portion of CH

you can get see the things latitude

longitude portion of Chennai directly

this L to longitude connect with this

website it will get the real time data

based on the realtime data it will

perform the prediction operation see the

thing so chenai belongs to Safe category

according to our ml model we did not

detect any sign of potential flood so in

chai you don't find any flood data okay

so you can check the results also in

chenai the average temperature will be

85 par maximum temperature will be 93

par what's the wind be in J we are

getting 30 this is today results okay uh

you wind speed 13. 78 mph as well as

what's the cloud coverage 88% of cloud

coverage is there okay what's the

humility ratio 72 these results are

belong to chinai chinai did not find any

flood things okay any potential flood

related things what I'm doing means

check the today date I'm just going for

the Delhi you can type Delhi here so for

Delhi also you can get the latitude

longitude so after getting the latitude

longitude it will collect all the real

time data with the help of API it will

get collect all the real time data

satellite maps and all then it will

predict how much maybe FL happening or

not you can get the result for Del

unsafe according to our ml model this

area has potential risk of flooding

within next 15 days potential risk of

landslide Also may happen please take

emergency measures this result for Del

okay depend upon the city the results

may change belongs to our machine

learning model you can get temperature

results per daily maximum temperature

minimum temperature as well as wind

speed cloud coverage then how much M uh

uh participants means the people R for

people ratio then how much humidity

happening okay sorry I mispronounced

that precipitation precipitation how

much percipitation will happen all the

ratios will get real time okay with the

help of this API after getting this API

after getting this data it will perform

the machine learning operation then only

it will give the result what I'm going

to do means I'm going to do for the uh

Mumbai Mumbai I'm giving results so if

I'm typing Mumbai

means so after typing Mumbai you can see

the results go for the things so this is

our Mumbai latitude longitude of Mumbai

it will get the all the datas from the

API then it will perform the machine

learning operation Mumbai FL prediction

safe okay according to our ml model we

did not find any sign of potential flare

what's the Mumbai temperature maximum

Wing speed precipitation value

everything you can see here then what

I'm going to use means I'm going to

enter Hyderabad also

here so I'm going to use hydrabad

I'm going to take

kraat so you can get the result for

Hyderabad also in our API so we typed

hyat this is the geological position of

Hyderabad then you can see the result so

safe usually in Hyderabad flight may may

not happen so we got the results also so

we did not find out any potential flood

okay this is Hyderabad result this

project not only belongs to Indian city

or Indian State you can get results for

any city in the world so I'm using

London here so London New York you can

type any City it will work

okay I'm just typing London so after

typing London so let me check it got

London or not okay you can see response

London latitude longitude so it got

result for London

also safe according to our M model we

did not sign any potential flood so if

we have checked I Delhi only having

potential risk of flood for today date

for any date this project will work

perfectly okay so London maximum

temperature average temperature wind

speed cloud coverage then precipitation

value humidity value so these are the

the results for London okay this is

completely end to end project with the

help of machine learning there is this

very interesting project as well as very

important for project for current days

okay this will help us to save millions

of lives if we applied perfectly to know

the accuracy of this project just go to

the accuracy tab you can get the

complete accuracy of the project so this

is accuracy score of our project you can

see the accuracy around

97% so this is our accuracy we got

around 97% accuracy for our propos to

system also what are the features we are

considered for training means so

drainage value Dam quality River

Management political factors then

population score these are the important

features we have considered to uh

execute training as well as testing of

the Project based on that these features

only it will DCT whether the flood may

happen or not we have consider drainage

also here see the things this is

confusion Matrix of our project then

overall architecture of our project so

this is overall architecture okay to get

this project please visit I expert we

have displayed various projects for best

price okay you can get all the projects

from this website okay we have displayed

various project on Mission learning as

well as blockchain thank you

[Music]