A STUDY ON USAGE OF DATA ANALYTICS IN AGRICULTURE FOR

CROP PREDICTION
Our Aim :-

1. Introduction :- 20 Pages - Done - Need to Check Plagarism

2. Review of Literature:- 20 Pages - Done - Need to do References
3. The Company / Organisation:- 20 Pages - Not Done Anything
4. Data Analysis and Inferences:- 30 Pages
5. Findings – 2 Pages
Summary – 2 Pages
Conclusion – 1 Page
6. References
7. Bibiliography

1
Title:- A STUDY ON USAGE OF DATA ANALYTICS IN AGRICULTURE FOR
CROP PREDICTION

Objective:-
The following the Objectives of the Project
1. To study the role of Data Analytics application in Agriculture.

2. To examine the advantages of using Data Analytics Techniques for Crop Prediction.
3. To develop a Machine Learning Model for crop prediction by using Python Coding and various
Statistical Concepts.

4. To predict the best crop to be grown based on Soil Composition, Temperature, Ph- Value and
Rainfall.

2
 INTRODUCTION

Data Analytics ? :-

Different application of Data Analytics

Application of Data Analytics in Agriculture

Business Analytics is the process by which businesses use statistical methods and
technologies for analysing historical data in order to gain new insight and improve
strategic decision-making. It is a data management solution refers to the use of
methodologies such as data mining, predictive analytics, and statistical analysis in order
to analyze and transform data into useful information, identify and anticipate trends and
outcomes, and ultimately make smarter, data-driven business decisions.

History…

COMPONENTS OF BUSINESS ANALYTICS:-

a) DATA AGGREGATION: prior to analysis, data must first be gathered,

organized, and filtered, either through volunteered data or transactional records.

b) DATA MINING: Data mining for business analytics sorts through large datasets
using databases, statistics, and machine learning to identify trends and establish
relationships.

c) ASSOCIATION AND SEQUENCE INDENTIFICATION: the

identification of predictable actions that are performed in association with other actions
or sequentially

d) TEXT MINING: explores and organizes large, unstructured text datasets for the
purpose of qualitative and quantitative analysis

e) FORECASTING: analyses historical data from a specific period in order to make

informed estimates that are predictive in determining future events or behaviours.

f) PREDICTIVE ANALYTICS: predictive business analytics uses a variety of

3
 statistical techniques to create predictive models, which extract information from
datasets, identify patterns, and provide a predictive score for an array of organizational
outcomes

g) OPTIMIZATION: once trends have been identified and predictions have been
made, businesses can engage simulation techniques to test out best-case scenarios

h) DATA VISUALIZATION: provides visual representations such as charts and

graphs for easy and quick data analysis.

Data Science in the Indian Agriculture Industry

Agriculture is the backbone of the Indian economy, but the industry currently requires more
support and concentration than any other. India is a country of over a billion people in
population. Around 70% of the population lives in the rural areas. With 40% of the country’s
workforce, agriculture is a major industry and an influencer of the Indian economy. Despite
this all facts, its contribution to the $2.3 trillion economy is just a meager 16% of the entire
GDP.

4
The reason behind this low contribution of Agriculture towards GDP of this country is many.
Agriculture in India lacks institutional attention, proper support from banks in terms of loans
and supportive farmer welfare schemes from the Government side. Sometimes, it even suffers
from a myriad of disasters like depleting groundwater levels in rural areas, climate change,
unpredictable monsoon or lack of rainfall, droughts, floods, unfair price fixing policies of
produce, migration of farmers towards the cities in search of better paying jobs, and more.

Agriculture is one sector responsible for feeding every individual, but the people involved in it
are not taken care properly. After many institutions have failed, time has indeed come for
technology to take care of the change. With newer problems cropping up every day in the most
inevitable indigenous sectors, it is high time we resort to emerging technologies for solutions.

Big Data to the Rescue

The revolutionary technology that goes by the name “Big Data” has already made impact in
other Indian industries from IT to healthcare. And now, investors and market players are
planning to leverage the potential of Big Data for the benefit of agriculture sector in India.
Apart from major companies, it is the vision of several youth of the country that has attracted
the use of Big Data for farming. For instance, SatSure, founded by the 33-year-old Abhishek
Raju works on using Big Data and its allied technologies like data science and IoT to better the
lives of farmers.

Abhishek shares the fact that he was deeply moved by the rate of farmer suicides and the lack
of application of science and newer technologies in the oldest Indian industry – agriculture. His
solution to this is SatSure. According to him, “The parameters associated with soil health and
crop growth have had a very restricted scope for research and his technology immensely uses
Big Data and Machine Learning technologies to solve the restriction and bring about insights
on crop phenology.”

When we got on a call with Mr. Abhishek Raju, he shared that, “Indian agriculture sector is
much disorganized and heavily cash oriented. Electronic transactions are almost non-existent,
and that is why most of the transactions are unrecorded. We help them provide insights about
farm productivity, when to irrigate, sow, harvest, and the patch of land that can be used by
farmers. We help banking and insurance companies in settlement of risk assessment, crop loss,

5
and offer insights by analyzing current and historical satellite images.”

“The satellite images are not only in a single visual spectrum but we have multiple data layers
which contain images merged into one to gather as much information as possible. This is what
we at SatSure mine. However, data is one thing, and what you infer from data is another thing.
We analyze data to make action oriented conclusion-able intelligence.” He adds.

Facilities like satellite-based filed monitoring, embedded sensors on crops and fields,
predictions on wind direction, fertilizer requirement notifications, pest infestations, GPS-
enabled tractors, water cycles, and more are acting as points of rich data sources that could be
used for better agriculture practices. Besides, Big Data and analytics now also enable
monitoring and supervision for growth rate and nutrient requirements on a plant-by-plant basis.
Moreover, analytics is enabling farmers to make data-based decisions like which crops to plant
for their next harvest. The rich information on soil health, water availability, and predictions on
rainfall and precipitation make this data source. Welcome to the world of Smart Farming.

Smart Farming

Smart Farming is the breakthrough application of science and technology in the field of
agriculture. Smart farming is the application of technologies like IoT, Big Data and analytics
on an agricultural field. It makes use of technologies like the Internet of Things, cloud
computing, Machine Learning, and Big Data to enable farmers to have more insights on the
consequences of their actions and take a much better and informed decision on farming
practices.

The power of smart farming lies in the fact that it goes beyond solving the shortcomings and
pitfalls of agriculture. The application of Big Data is leaving significant impact on the entire
realm of supply-chain, giving predictive insights on farming practices and operations, help
redesign business models, deliver realtime decisions on operations and more.

Jyoti Vaddi of Cropin shares, “The world population is estimated to cross the 10 billion mark
6
by the middle of this century. This population growth combined with urbanization will require
the agricultural production to double. To succeed, Jyoti recommends the need for smart
solutions for fairly produced, sustainable food, feed, and fibre, which is one of the mainstay
principles of CropIn.”

Satsure enables insurance companies, banks, traders, pesticide and seed manufacturing
companies, and farmers the ability to take informed farming decisions by leveraging the
combined potentials of technologies like –

 Cloud Computing
 Big data
 Machine Learning
 The Internet of Things
 Web-based Software as a Service platforms

Raman Singh Saluja, founder  of Gramco Infratech says, “Agriculture is a very physical
business, requiring physicality in terms of handling/warehousing/value addition / etc, which
will continue to be the bedrock.”

When it comes to the use of technologies, he shares that data analytics offers tremendous
potential for improving cost to output ratio, reduce/optimize Input usage, increase yields, offer
timely actionable information and do more.

He further reveals,” At Gramco, there are two initiative underway which will be brought to
market by the 3-4th quarter of 2018. One has been piloted with very encouraging results with
on ground support of a leading insurance company.”

7
During their market opportunity study, Cropin gathered that agribusinesses had minimal and
outdated technological/digital resources, and were not able to make informed data-driven
decisions. She reveals that with consumers keen to know the origin of their food and how it
was produced and processed, there was a need for transparency along the end-to-end
agribusiness supply chain.

The technologies that power Cropin in delivering efficient farming solutions to stakeholders in
its network include:-

 The use of app-based data generation and extraction

 Data storage on the cloud
 Satellite Monitoring
 Machine Learning and Real-Time data visualization

These technologies foster an environment for production forecast, risk management and
coverage, output predictability, quality maximization, and increased farm sustainability to
agriculture input companies, banks and financial institutions, insurance companies, farming
enterprises, seed manufacturing companies and government bodies respectively.

Components of Smart Farming

Smart farming is a network of interdisciplinary and complementing technologies and facilities.

8
The components of smart farming are best if they comprise of the following:

Management Information Systems

This is generally the database where all chunks of data from multiple sensors and resources are
gathered, stored, analyzed, and retrieved for actions. An optimized management information
system should offer information on:

Crops:

 Crop stress
 Statuses on crop tissue nutrients
 Crop population
 Weed patches
 Fungal or insect infestation
 Crop yield

 Soil

 Physical condition
 Soil texture
 Structure
 Moisture
 Nutrients and more

 Climate

 Humidity
 Rainfall
 Wind speed
 Temperature

 Devices

Technology is what puts the smart in smart farming and the following make up the network:
9
  Global positioning systems and differential global positioning systems for better
accuracy
 Geographical information systems
 Remote sensing technologies like data sensors, RADARS, data transmitters, drones,
cameras, and other connected devices
 Cloud architecture
 The Internet of Things, where devices are capable of communicating with each other
and deliver real-time updates and notifications to farmers on crop statuses, water levels,
moisture content, crop yield, and more.Technologies like Machine Learning, Data
Analytics, and Big Data for the entire process and setup to make sense.

What is Precision Agriculture

Also referred to as Site-specific Crop Management System or Satellite Farming, this is a

concept in farming that relies on observation, measurement, and response to various inbound
and outbound requirements in agricultural fields.

The primary vision of precision farming is to optimize RoI and preserve resources by allowing
farmers and landlords to take optimized and informed decisions from the available field data.
Precision agriculture fosters an environment where farmers can zero-down precise locations in
their fields for the spatial availability of several resources like water availability, topography,
soil fertility, organic matter, nitrogen levels, moisture content, the presence of magnesium,
potassium, and more.

Complemented by services and features like GPS devices, sensors that are even capable of
measuring chlorophyll levels, drones, and satellite imagery, precision agriculture offers a
treasure chest of information for farmers.

Big Data Applications in Agriculture: Role, Importance & Challenges

Big data is an extensive collection of both structured and unstructured data that can be mined
for information and analyzed to build predictive systems for better decision making. Besides
the government, telecom, healthcare, marketing, education, and several industrial sectors, big
data applications in agriculture are gaining momentum as technologies like livestock
monitoring gadgets, drones, and soil sensors are generating large volumes of data to support
data-driven farming. The ultimate goal is to help farmers, agriculturists, and scientists adopt
1
0
beneficial farming practices.

Big data 

When a question arises what is big data, it is a collection of large, complex, and unprocessed
data is called ‘big data’. Due to complexity, big data cannot be processed by conventional data
processing and data management applications and requires advanced tools that can analyze and
process large volumes of data. Big data is characterized by some unique features – volume,
variety, velocity, variability, veracity, and complexity. This vast reservoir of information must
be studied, stored, and processed systematically for its applications in the public sector,
scientific research, agriculture, industry, etc.

Applications of big data 

Government – Data influx from sources such as sensors, satellites, CCTV and traffic cameras,
calls, emails, social media, IT spaces, academia, etc. calls for efficient data storage and analysis
for better governance and management of the public sector.

Banking – The big data applications in banking & insurance sector handles enormous amounts
of data. Big data analytics are being used to store data, improve scalability, and derive business
insights.

Healthcare – The problem of communication silos that plagues the healthcare industry can be
considerably reduced with the application of big data-based protocols.

Telecom – Real-time analysis of big data provides useful predictions to derive business
insights and strategies such as delivering revenue-generating services while keeping in mind
network and customer considerations.

Big data in agriculture 

Big data applications in agriculture are a combination of technology and analytics. It entails

the collection, compilation, and timely processing of new data to help scientists and farmers
make better and more informed decisions. Farming processes are increasingly becoming data-
enabled and data-driven, thanks to smart machines and sensors that generate vast amounts of
farm data.

1
1
Traditional tools are being replaced by sensor-equipped machines that can collect data from
their environments to control their behavior – such as thermostats for temperature regulation or
algorithms for implementing crop protection strategies. Technology, combined with external
big data sources like weather data, market data, or standards with other farms, is contributing to
the rapid development of smart farming. 

Role of big data in agriculture 

Sustainability, global food security, safety, and improved efficiency are some of the critical
issues that are being addressed by big data applications in agriculture. Undoubtedly, these
global issues have extended the scope of big data beyond farming and now cover the entire
food supply chain. With the development of the Internet of Things, various components of
agriculture and the supply chain are wirelessly connected, generating data that is accessible in
real-time.

Primary sources of data include operations, transactions, and images and videos captured by
sensors and robots. However, extracting the full potential of this data repertoire lies in efficient
analytics. The development of applications related to risk management, sensor deployment,
predictive modeling, and benchmarking, has been possible due to big data. 

Stakeholders involved 

Technology and input suppliers are the traditional players who offer their platforms and
solutions to the farmers. Data privacy and security risks compel farmers to form coalitions to
benefit from their data, creating a close and proprietary environment. Big data also attract start-
ups, private firms, non-agricultural tech companies, and public institutions.

The organization of the stakeholders determines the infrastructure of big data solutions – either
proprietary or an open-source system. The development of big data applications in
agriculture will result in either the farmers becoming franchisers in integrated long supply
chains or a scenario in which farmers collaborate with suppliers and the government to engage
in short supply chains.

How is big data analytics transforming agriculture? 

Boosting productivity – Data collected from GPS-equipped tractors, soil sensors, and other
external sources has helped in better management of seeds, pesticides, and fertilizers while
1
2
increasing productivity to feed the ever-increasing global population.

Access to plant genome information – This has allowed the development of useful agronomic
traits.

Predicting yields – Mathematical models and machine learning are used to collate and analyze
data obtained from yield, chemicals, weather, and biomass index. The use of sensors for data
collection reduces erroneous manual work and provides useful insights on yield prediction.

Risk management– Data-driven farming has mitigated crop failures arising due to changing
weather patterns.

Food safety – Collection of data relating to temperature, humidity, and chemicals, lowers the
risk of food spoilage by early detection of microbes and other contaminants.

Savings – AI and data analytics-driven farming generate significant savings for the agriculture
industry.

Big data analytics has influenced some of the most critical sectors of the economy and will
continue to do so. The big data applications in agriculture are still in their early days, with
challenges that need to be addressed. The full potential of big data will be realized if farmers
and stakeholders come together to develop and adopt innovative crop management techniques
that are data-driven and data-enabled.

If you are interested to know more about Big Data, check out our , which is designed for
working professionals and provides 7+ case studies & projects, covers 14 programming
languages & tools, practical hands-on workshops, more than 400 hours of rigorous learning &
job placement assistance with top firms.

Data Science in Agriculture – Advancing Together & Benefiting Farmers

Data Science in Agriculture – Saving the lives of farmers

Agriculture is the most important sector of the Indian economy, that provides employment to
almost half the population of the country’s workforce!! India is the second-largest producer of
fruits and vegetables in the world.

Agriculture is the backbone of Indian economy, still, it suffers from a heaping number of
disasters such as climate change, unpredictable monsoon or lack of it, droughts, floods,
1
3
migration of farmers towards the cities in search of better-paying jobs, and more. People
involved in agriculture are the last to be taken care of, even when they are the one who feeds
the whole country. With institutions failing to support agriculture in terms of providing loans
and farmer welfare schemes, the time has come for technology to take over the change. Data
Science is here for the rescue!

Data is the necessity for industries and hence, data science has a number of applications. After
revolutionizing industries like IT, Banking, Manufacture, Finance, Healthcare, and many more,
it is all set to benefit the agriculture industry.

Data Science Applications in Agriculture:-

Here are the six applications of data science in agriculture sector:

1. Digital Soil and Crop Mapping

This is related to building digital maps for soil types and properties. Some people in the
agricultural industry manage so many acres of land, it’s almost impossible to get prompt
updates and alerts about potential problems without help from technology.

Farmers Edge, a Canadian company takes daily satellite images of farms and combines it
1
4
with other relevant data. It includes information from more than 4000 interconnected weather
stations!
Several countries like Ireland also depend on satellite-based soil and crop monitoring to inspect
areas more quickly than traditional methods allow. This helps in deciding what crops should be
grown on a particular piece of land. It saves a lot of time and effort and results in higher yield
production.

2. Weather Prediction

Weather plays a very important role in agriculture production and has an influence on the
growth, development, and yield of crops. Weather aberrations can cause physical damage to
crops and soil erosion.

The quality of crops from the field to the market depends on the weather. Bad weather can
adversely affect the quality of crop during transportation or storage.

Data science experts know how to use tools that identify the patterns and relationships that may
otherwise be hidden. They can draw conclusions that push agricultural science forward through
the examination of specific factors leading to change in weather.

The findings brought about by sifting through databases and studies to conclude things like this
in agricultural processes can bring about remarkable changes.

Elements of agricultural weather forecasts are:

 Amount and type of coverage of sky by clouds
 Rainfall and snow
 Max, min, and dew point temperatures
 Relative humidity
 Wind speed and direction
 Low-pressure areas, cyclones, tornadoes, and depressions
 Events like fog, frost, hail, thunderstorms, and wind squalls

3. Fertilizers Recommendation

Knowing the exact fertilizer rate is a science and requires a thorough analysis of multiple
factors. Often, hundreds of dynamic parameters have to be considered.

1
5
Such parameters include crop nutrient uptake rates, research data, soil chemical, physical and
biological properties, weather, water composition, land type, soil testing methods, irrigation
techniques, fertilizer characteristics, interactions between fertilizers and many more.

Because of the complexity of finding the “optimal fertilization range”, misuse of fertilizers is a
global phenomenon. The majority of farmers still rely on trial and error, guesswork and
estimation.

The result is, crops do not meet their yield potential, and increase environmental pollution.
Data science professionals are now able to advise the farmers with the right quantity of
fertilizers.

4. Disease Detection and Pest Management

In modern agriculture, advanced algorithms are used to identify the patterns and behavior of
nature that helps in forecasting the invasion of pests and the spread of microscopic diseases.

Advanced analytics in agriculture are informing how farmers should manage pests. Digital
tools and data analysis in agriculture are being utilized to scientifically deal with harmful
insects.

Agricultural pests can quickly cut into a farmer’s profits. But, misusing pesticides can have
adverse effects on people, plants and other living things.

Fortunately, some companies have recruited data science professionals to develop user-facing
platforms that analyze when to apply pesticides and how much to use.

While some insects can be incredibly beneficial to farmers and the crops, others can be toxic
and spread diseases. Disease detection can be done by taking images of the field using drones
and processing them to detect areas within this field that are infected.

5. Adaptation to Climate Change

Climate change is a looming concern that has already affected the agriculture sector. However,
data science experts are working hard to figure out ways to compensate for the change.

One project involves giving IoT sensors to Taiwanese farmers of rice production so they can
collect information that is necessary about their crops. It will help farmers to optimize their
production cycles, even if climatic changes make it challenging.

1
6
The traditional farming calendar is no longer sufficient due to the extreme climate changes but,
data analysis can revolutionize the future of farming.

Data scientists are also analyzing agriculture soil data to understand how soil can cope with
climate change by releasing greenhouse gases and also how soil can adapt to climate change.

6. Automated Irrigation System

We can also use weather prediction in the automated irrigation system. How exactly? Having a
system that tells you the soil is dry but you don’t need to irrigate because it is going to rain
after a few hours seems to be an interesting use case.

All the countries in the world are currently in a situation where they are required to use water in
a very efficient manner.

According to the recent studies, water is becoming more and more in short supply worldwide
and more than one-third of the world population would face total water shortage by the year
2025.
In agriculture too, the major problem which farmers face is the water scarcity, so to improve
the usage of water, one of the irrigation systems- using drip irrigation which is implemented as
automated irrigation system for small scale farms.

Data Science Case Study on DATOS Project

According to a recent press, DATOS project applied artificial intelligence, machine learning,
and other data science techniques to remotely sensed data. Apart from that, it used systems to
produce geospatial outputs that can be used for disasters, agriculture and other purposes.

The DATOS Project has developed a way to map out crops by using satellite images and by
extracting the temporal signature of crops determined through radar satellite images.

Moreover, DATOS produces flood situation maps by retrieving satellite images and letting AI
identify flooded areas from these images.

It is able to detect floods in areas that are hit by heavy rainfall and sends these mapped out
areas to the respective DOST regional offices in the event of severe weather disruptions.

1
7
Data science is also used to detect objects from satellite images. An example is the current
partnership with the Bataan Peninsula State University in order to help the institution automate
the mapping of their province’s mango trees.

Other objects that have been detected from satellite images include road networks, ships, land
cover classes, and built-up areas.

Real-life Examples of Data Science in Farming

 A good example is in Egypt, where farmers use water pumps to collect water from the
river Nile to their crops. Advanced water sprinklers are used to irrigate large fields and this
helps the crops to get enough water which is essential in their growth. This method is
useful in the efficient usage of water.
 The invention of the mobile app such as “FamGraze” for farmers manage their grass
more effectively by suggesting the cheapest feed for their livestock. It saves time and there
is no need for any paper or spreadsheets. This helps in working faster and being accurate in
and off the field!
 Modern transportation technology facilities help farmers easily transport fertilizers or
other farm products to different parts of the world. For example – consumers in Dubai will
get fresh carrots from Africa on the same day when the carrots leave the garden in Africa.

Summary

These were just a few cases showing how one can use data science in the work of an
agricultural corporation right now because, in the future, we cannot predict what opportunities
it can bring. Technology has played a huge role in developing this industry. Today, it is
possible to grow crops in a desert by use of agricultural biotechnology and there is much more
scope in the future.

Data scientists are part of virtually every major industry, and agriculture is no exception. Here
are six ways data scientists bring advantages to the agricultural sector.

1. Receiving Useful Data to Help Fight Food Scarcity and Empower Small Farmers

1
8
Since data scientists have tools to process and analyze gigantic amounts of data efficiently,
projects are underway to determine how that information might help small-scale farmers join in
the battle to solve worldwide food shortages.

In September 2018, a coalition launched a project that will run through 2030 and look at data
from approximately 500 million farmers in impoverished areas from 50 countries.

The people behind the project hope the data will show whether agricultural investments in
various countries are paying off and help develop policies for the farmers. On a larger scale, this
project aligns with the United Nations’ Sustainable Development Goals to double the
agricultural productivity and incomes of farmers in developing nations and help them reduce
world hunger.

2. Managing Crop Diseases and Pests

Agricultural pests can quickly cut into a farmer’s profits. But, misusing pesticides can have
adverse effects on people, plants and other living things. Fortunately, some companies recruit
data scientists to help them develop user-facing platforms that analyze when to apply pesticides
and how much to use.

One of them is a Brazillian company called Agrosmart. Its technology relies on Internet of

Things (IoT) sensors and artificial intelligence to determine the kind of insects on a crop and the
quantity present. Farmers then get an associated report and can use it to plan their pest
management approaches. The goal is to help farmers cost-effectively control pests with a
minimized environmental impact.

In another case, Saillog, an Israeli startup, developed a smartphone app called Agrio that
informs farmers of the diseases currently affecting their crops or ones found on surrounding
farms.

3. Investigating Agricultural Niches

Data scientists know how to use tools that identify patterns and relationships that may otherwise
remain hidden. As such, they can draw conclusions that push agricultural science forward
through the examination of specific factors. For example, researchers know trace minerals
1
9
positively affect the metabolic functions of livestock and poultry, while carotenoids play a role
in increasing egg yolk quality and nutrition.

The findings brought about by sifting through databases and studies to conclude things like
these show how seemingly small factors in agricultural processes can bring about substantial
changes. When animal feed brands, farmers and others in the agriculture sector take advantage
of such findings from data scientists, they could positively change their operations and get better
results.

4. To Cope With Climate Change

Climate change is a looming concern that has already affected the agriculture sector. However,
data scientists are hard at work figuring out ways to compensate for the shift.

One project involves giving IoT sensors to Taiwanese rice farmers so they can collect crucial
information about their crops. It’ll all go into a database used to help farmers optimize their
production cycles, even when climate change makes that task exceptionally challenging.
Following the traditional farming calendar is no longer sufficient because of climate change.
But, data analysis could forever change the future of farming.

Scientists are also scrutinizing agricultural soil data to improve their understanding of how soil
contributes to climate change by releasing greenhouse gases, as well as how soil data might aid
in adapting to climate change. Collecting this kind of information is tricky, but scientists believe
it could fill in knowledge gaps associated with the relationship betweensoil and climate change.

5. To Make Yield Predictions

A poor yield can result in a devastating season for farmers, as well as all the entities that depend
on the crops. IBM has a platform that estimates corn yields two to three months in advance,
reducing unpleasant surprises for agricultural professionals.

Similarly, researchers at the University of Illinois depend on seasonal forecasts and satellite data
to make end-of-season predictions earlier than usual. Lab results indicate this new method is
even more accurate than the U.S. Department of Agriculture’s real-time data.
2
0
6. To Provide Images of Crops and Land

Some people in the agricultural industry manage so many acres of land, it’s impossible to get
prompt updates and alerts about potential problems without help from technology. Canadian
company Farmers Edge takes daily satellite images of farms and combines it with other relevant
data, including information from more than 4,000 connected weather stations.

Moreover, several countries, including Ireland and Ghana, depend on satellite-based crop
monitoring to inspect areas more quickly than traditional methods allow.

Data scientists play an essential role in furthering the use of satellite imagery for these purposes
and others. Pilot projects and new platforms could help those experts determine the methods
most appropriate for tackling current and future needs within the agriculture sector.

An Exciting Time for Data Scientists Interested in Agriculture

Agriculture has become increasingly high-tech over the years. Now, much more than in the past,
agriculture professionals can dig into data and use it to make highly informed decisions. The
advancements of data scientists are making this reality possible, both now and for the
foreseeable future.

PROFILES:

INDUSTRY PROFILE

Agro-based industries are industries that use plant and animal-based agricultural output
2
1
as their raw material. Also, they add value to agricultural output by processing and
producing marketable and usable products. Some examples of agro-based industries in
India include Textile, Sugar, Vegetable Oil, Tea, Coffee and Leather goods industries.

Importance of Agro-based industries in India

All branches of agro-based industry are important because:
a) Help in increasing industrial production.
b) Provides employment to landless agricultural labour and tribal population from rural
and backward areas.
c) Ensure the development and stability of rural economy through diversification and
reduced dependence on agriculture.
d) Ensure the alleviation of poverty by providing steady sources of income and
livelihood.
e) Earn much required foreign exchange for the country.
f) Improve the standard of living in rural areas.
g) Help in reducing the extreme inequalities in the distribution of income and wealth.
h) Are easy to establish.

i) Support balanced growth between agriculture and industry, and

j) Helps in avoiding wastage of perishable agriculture.

2
2
REVIEW OF LITERATURE

There is lot of research done on “Crop Prediction” based on soil health,

average humidity, etc. but there are five relatively recent studies that have
received a lot of attention:-

1. Pavan Patil, Virendra Panpatil, Professor Shrikant Kokate[2020]

Decision tree shows poor performance when dataset is having more variations
but naïve bayes provides better result than decision tree for such datasets.
The combination classification algorithm like naïve bayes and decision tree
classifier are better performing than use of single classifier model.

2. S.Veenadhari, Dr Bharat Misra, Dr CD Singh [2019]

This research is to study the various Machine learning approaches for
forecasting crop yield based on climatic parameters. This study has reported a
correlation between usage of chemicals and crop yield.

3. Viviliya.B and Vaidhehi.V [2019]

The attributes in the dataset are the soil type, groundwater level, rainfall,
water availability, temperature of one dataset and the other dataset included
the potassium, phosphorus, and nitrogen values, fertilizers, soil pH and
organic carbon value. The dataset was preprocessed using basic preprocessing
tasks. Naive Bayes and J48 classifiers were used for the crop
recommendation. The final recommendation was done using association rules
based on the results obtained from the classifiers. The model was trained
using 10-cross validation. The testing was done based on different metrics
like the Accuracy, ROC Area, Recall, Precision, F-Measure etc.

4. D. Team, DataFlair, 27-Sep- [2019].

Though, the field of agriculture is less impacted by the recent technological
advancements. However, agriculturalists are rapidly moving towards working
with modern tools and technologies. One such up to the minute technology is
Big Data analytics. Big data has been introduced to almost every other sector
even agriculture is not outdistanced from it.

5. M. N. Islam Sarker, M. Wu, B. Chanthamith, S. Yusufzada, D. Li, and J.

Zhang [2019]
“Big data analytics which is often the complex process of examining large
and varied data sets, or big data, to uncover information – such as hidden
patterns, unknown correlations, market trends and customer preferences
which helps organizations make informed business decisions”.

6. R. V. McCarthy, M. M. McCarthy, W. Ceccucci, and L. Halawi [2019]

It is a technique used in big data analytics which can predict the future
outcome based on gathered or historic data. It is one of the very few
technologies that have the potential to remold any business positively. It
majorly consists of three important techniques also called as the “Big 3”
techniques: which are regression analysis, decision trees, and neural networks.
7. M.Peixeiro, Medium, 07-Aug-[2019].

In a time series analytic model, formerly observed values are used to forecast the output.It
uses the concept of time series which is a series of data points arranged in time. However,
here time is considered as an independent variable with an ambition to forecast the future. It
is used in forecasting crops, vegetation price movement and price fluctuation in market

8. Igor Oliveira, Renato L. F. Cunha, Bruno Silva, Marco A. S. Netto[2018]

This research gave insights to users with the capability to perform strategy
changes, like choosing more robust genetic variation before planting or even
changing the crop type, in order to accommodate for extreme climatic
variations further ahead in the crop cycle. This research focuses on Scalable
Machine Learning System for Pre-Season Agriculture Yield Forecast.

9. Olakunie Elijah, Tharek Abdul Rahman, Igbafe Orikumhia [2018]

The surge in global population is compelling a shift toward smart agriculture
practices. This coupled with the diminishing natural resources, limited
availability of arable land, increase in unpredictable weather conditions
makes food security a major concern for most countries.

10. Shivi Sharma, Geetanjali Rathee and Hemraj Saini [2018]

In this work, soil and environment features i.e. average temperature, average
humidity, total rainfall and production yield are used in predicting two
classes namely: good yield and bad yield.

11. Barun Basnet, Junho Bang [2018]

In this study it tells the use of data analytics has shifted agriculture from
input- intensive to knowledge-intensive as a large amount of agricultural
data can be stored, shared, and analyzed to create information.
12. Z. Doshi, S. Nadkarni, R. Agrawal and N. Shah [2018]
The accuracy of a machine learning algorithm may depend on the number of
parameters used and to the extent of correctness of the dataset. This research
focuses on Agro Consultant: Intelligent Crop Recommendation System Using
Machine Learning Algorithms.

13. K. Charvat [2018]

This research tells “An accurate understanding from the huge amount of data
can be interpreted in the form of overviews, summary, verifiable models etc.
However, most of the interpretation and visualization is organized in an Ad-
hoc manner. The models are basically visualized in the form of tables, graphs,
spreadsheets etc’.

14. P.Kumar, A.Kumar, S.Panwar, S.Dash, K.Sinha , V.K.Chaudhary,

M.Ray, [2018].

A complex and massive collection of data which is hard to process by

applying traditional data processing techniques or on-hand database
management tools is referred as ‘big data”. It reveals agriculture is highly
 depended on population and weather which produces data drastically. To
handle data, tools and techniques in big data are introduced.
15. Ch. Chandra Sekhar, J. UdayKumar, B. Kishor Kumar, Ch. Sekhar,
[2018].

The semi-structured is data that are able to convert from unstructured to

structured using available descriptions (xml document).In big data analysis,
raw format data are converted into standard format with the support of tools.
This is a process of information gathering, data analysis, visualization and
scheduling. In big data analysis, raw format data are converted into standard
format with the support of tools.

16. T. Edwin Nov. 13 [2017]

Our country had undergone several fluctuations in the price of onions last
year. The price of the onions increased from Rs. 26 per kg to Rs. 50 per kg in
the month of August. So most of the farmers decided to cultivate onion in
their fields seeing this huge increase in price so that they could make large
profits from their land.

17. Md. Tahmid Shakoor, Karishma Rahman, Sumaiya Nasrin Rayta,

Amitabha Chakrabarty [2017]
This paper helped us for selecting various attributes like land capability
classification, soil depth, slope, drainage, texture, erosion and permeability.
This research focuses on Agricultural Production Output prediction Using
Supervised Machine Learning (SML) Techniques.

18. Mayura Nagar, Mukesh Kumar [2017]

when weather is and rain come late or less farming decrees directly I ratio
but when farmer produce more crops then our market and storage is not
prepare for holding that due to this farmer force to sell his production in
minimum rate and face loss in farming.

19. S. Rajeswari, K. Suthendran; K. Raja Kumar [2017]

Cloud based big data analytics and the IoT technology performs an important
role in the feasibility study of smart agriculture. Smart or precision
agricultural systems are estimated to play an essential role in improving
agriculture activities.

20. Jharna Majumdar, Sneha Naraseeyappa, Shilpa Ankalaki [2017].

The findings of this study are “The estimation of production of crop helps
these companies in planning supply chain decision like production
scheduling. The industries such as fertilizers, seed, agrochemicals and
agricultural machinery plan production and activities like marketing based on
the estimates of crop yield”.
 21. Dhivya B H, Manjula R, Siva Bharathi S, Madhumathi . R [2017]

The following is the outcome of the research, Farmers also grow non-food
items like rubber, cotton, jute etc. More than 70% of the household in the
rural area depend on agriculture. This domain provides employment to more
than 60% of the total population and has a contribution to GDP also (about
17%).

22. S. Rajeswari, K. Suthendran, and K. Rajakumar, [2017]

The major influence of implementing big data in agriculture are

benchmarking, analytics, model prediction, visualization, marketing and
management.It is also dependent on two major factors.

23. Swarupa Rani [2017]

This research has discussed the application of mathematical model like fuzzy
logic designs in optimization of the crop yield, artificial neural networks in
validation studies, genetic algorithms designs in accessing the fitness of the
model applied, decision trees, and support vector machines to study soil,
climate conditions and water regimes related to crop growth and pest
management in agriculture.

24.DakshayiniPatil, Dr. M .S Shirdhonkar [2017]

This surveys the technical achievements in the field of Rice crop yield prediction.
It also Discussed methodology, comprehensive survey of many proposed approaches to predict
rice crop yield and applications. It also discusses various data mining methods used for
prediction of crop yield for rice. Rising better plans to foresee crop productivity in various
climatic conditions can help farmer and different partners in essential basic leadership as far as
agronomy and product decision.

25. Mukesh kumar , Prof.Mayura Nagar [2017]

Precision agriculture gives not only great challenges but also great
opportunities for Computer Scientists specially who are working in the field
of data analysis. Most batch processing tools follow Apache Hadoop which is
an open-source software which is reliable and very fast. Also, Hadoop
framework has been used for taking decision on storage and distribution of
the products in an efficient way without wastage.

26. Francesc X.Prenafeta-Boldu, Andreas Kamilaris, Andreas Kartakoullis

vol. 143, no. C, p. 14 [2017].

This research has revealed that “Tools for better demand and yield prediction
should be developed”. Among the suggested future directions, agricultural robots which are
self-operated to identify weeds and remove them, tools with high performances, programming
language for big data analysis can be identified. It is revealed that data privacy is an issue in big
data analysis in agriculture because misusing of data.

27. D.P.Acharjya, Kauser AhmedP vol. 7, no. 2, p. 9, [2017].

This research has identified the growth of big data was doubled at the end of
2016. Also, big data is characterized by 3Vs to 4Vs which is meant by
Volume, Variety, Velocity and Veracity. Through this research, we came to
know that Map Reduce, Storm and Apache Spark are emerging tools in big
data analytics. Most batch processing tools follow Apache Hadoop which is
an open-source software which is reliable and very fast.

28. Van-Quyet Nguyen, Giang-Truong Nguyen, Sinh-Ngoc Nguyen, Jintae

Choi, Kyungbaek Kim [2017].

The framework Hadoop is having two major parts which are HDFS and
Hadoop MapReduce. Apache Spark is also a computing framework for real
time data as well as having memory oriented architecture with
flexibility.This research has also revealed that “Apache Hadoop and Apache
Spark most popular among big data analytics applications”. In order to
minimize the restrictions regarding data collecting, data storage and data
analyzing it is proposed to use deep learning techniques in near future.

29. S. Pudumalar, E. Ramanujam, R. H. Rajashree, C. Kavya, T. Kiruthika

and J. Nisha [2016]

The crops that were considered in the model for prediction include coriander,
pulses, cotton, paddy, sorghum, groundnut, sugarcane, banana and vegetables.
Different attributes of the soil were considered in order to predict the crop,
which included pH, depth, erosion, permeability, texture, drainage, dater
holding and soil color. The technique used was ensembling, which combined
the power of using two or more different models for better prediction. The
ensembling technique used was called the Majority Voting Technique.

30. T.R. Lekhaa [2016]

The paper hypothesizes analysis of Explorative Data and considers the design
of different types of predictive models. A data set is taken as a sample data
set, and different regression techniques are tried to recognise and examine
each property. Specific regression methods discussed here are Multiple
Linear, Linear, Non-Linear, Polynomial, Ridge regression and Logistic. Using
this article, we obtain a comparative study of the different algorithms in data
analytics. This helped in determining which algorithm is most appropriate to
the proposed system.

31. Majumdar J, Ankalaki S [2016]

 This research helps in Knowing what crops has been grown, and how much
area of it had been shown historically, combined with the prices at which it
could have been sold at the nearest market-place provides the income-growth
profile of the farmer.

32. R. Kumar, M. P. Singh, P. Kumar and J. P. Singh [2015]

The crops were inspected and graded depending on an examination to estimate crop
yielding. This categorisation is found from different data mining algorithms. This
paper provides a perception into various grouping rules, such as K-Nearest
Neighbour and Naive Bayes. By making use of this document, we evaluated the
classification rules and established which all will match the set of data we will be
using in our project.
Step-by-step Process :-

Step – 1:

a) Importing Pandas and Numpy for Manipulating the Dataset

b) Importing Matplotlib and Seaborn for Data Visualization.
c) Ipywidgets for Interactivity.

Step – 2:

a) Read the Dataset

b) Checking the shape of the Dataset.
c) Checking the Head of the Dataset.

Step – 3:

a) Checking whether the Dataset is having any missing values.

b) Summary of all crops
c) Check the Summary Statistics for each of the Crops i.e Minimum, Maximum and Mean
of each variable.
d) Compare the Average Requirement for each crops with average conditions

Step - 4:

a) the distribution of Agricultural Conditions

b) find out some Interesting Facts
c) Understand which crops can only be Grown in Summer Season, Winter Season and
Rainy Season

Step - 5: Clustering Similar Crops

a) Try to Cluster these Crops

b) Determine the Optimum Number of Clusters within the Dataset
c) implement the K Means algorithm to perform Clustering analysis

Step – 6: Visualizing the Hidden Patterns

a) Data Visualizations

Step -7: Predictive Modelling

a) Split the Dataset for Predictive Modelling

b) Create Training and Testing Sets for Validation of Results
c) Create a Predictive Model
d) Evaluate the Model Performance

Step – 8: Real time Prediction

a)
 Data Analysis and Inferences

The Dataset also consists of Soil Conditions required to grow the Crops. They are :-

 N: The Ratio of Nitrogen Content in the soil.

 P: The Ratio of Phosphorus Content in the soil.
 K: The Ratio of Potassium Content in the soil.
 pH: pH of the soil.

The Dataset consists of Climatic conditions required to grow the crops. They are :-
 Temperature
 Humidity
 Rainfall.

The Dataset consists of 22 unique crops such as Maize, Wheat, Mango, Watermelon, Rice,
Chickpea, Kidney beans, Pigeon peas, Moth beans etc.

Machine Learning Algorithm:-

a) Clustering
K Means algorithm to perform Clustering analysis
b)
PROBLEM STATEMENT

4. RESEARCH METHODOLY

Need for the study

Earlier, crops were sown based on farmer’s intuition, past knowledge. This is
trial- and-error method. There is a need for scientific, statistical based approach
while selecting a crop. The crop should be selected based on the fertility of the
soil, availability of Ground-water level and other factors. This approach will
increase the yield of crop, decrease the losses to farmers, increases profits etc.

The Objectives of the Study

The following are the Objectives of the Study:-

a) To study Business Analytics application in Agriculture.

b) To examine the advantages of using Business Analytics Techniques for
Crop Prediction.
c) To examine various Challenges related to Business analytics.
 d) To develop a Machine Learning model for predicting the best suitable
crop to based on the Soil Condition, Ground water Table and Seasoning
effects.

Scope of the Study

a) This study is confined only to understand the Advantages, Disadvantages

of Business Analytics for Agriculture, Crop Prediction.

b) The source of the Data is collected only from ‘Harvestify’, ‘GitHub’ and
‘Kaggle’.
c)
d) The crop is suggested based on the Soil Health, Ground water Table,
Seasoning effects etc. This study is confined only these four parameters.
 Sources of Data: -

This research is entirely based on Secondary Data.

Secondary Data:-The Data is collected from a Machine Learning based website

“Harvestify” and from ‘GitHub’ (is a provider of Internet hosting for Software
Development). Ground water level is collected from Kaggle Platform.

Sampling Techniques: -

For this study, convenient sampling is being used

a) Simple Linear Regression

b) Multiple Linear Regression
c) K- Means Clustering

Statistical Software’s:-
a) Anaconda Software
b) Jupyter Notebook
c) Python Coding tools are used to analyze the collected Data.
Youtube link :- https://www.youtube.com/watch?v=323jreAO8O0

Youtube :- https://www.analyticsvidhya.com/blog/2020/11/artificial-intelligence-in-
agriculture-using-modern-day-ai-to-solve-traditional-farming-problems/

REFERENCE
2019:-
1. Viviliya, B. and Vaidhehi, V., “The Design of Hybrid Crop Recommendation System
using Machine Learning Algorithms”. International Journal of Innovative Technology
and Exploring Engineering, 2019, 9(2), pp.4305-4311.

2. D. Team, “Data Science in Agriculture - Advancing Together and Benefiting Farmers,”

DataFlair, 27-Sep-2019. [Online]. Available: urlhttps://dataflair.training/blogs/data-
science-in-agriculture/. [Accessed:02-Dec-2019].

3. M. N. Islam Sarker, M. Wu, B. Chanthamith, S. Yusufzada, D. Li,and J. Zhang, “Big

Data Driven Smart Agriculture: Pathway for Sustainable Development,” in 2019 2nd
International Conference on Artificial Intelligence and Big Data (ICAIBD), 2019, pp.
60–65, doi:10.1109/ICAIBD.2019.8836982.
4. R. V. McCarthy, M. M. McCarthy, W. Ceccucci, and L. Halawi, “Introduction
to Predictive Analytics,” in Applying Predictive Analytics, Cham:Springer International
Publishing, 2019, pp. 1–25.

5.M. Peixeiro, “The Complete Guide to Time Series Analysis and Forecasting,” Medium, 07-
Aug-2019. [Online]. Available: https://towardsdatascience.com/the-complete-guide-to-time-
seriesanalysis-and-forecasting-70d476bfe775. [Accessed: 06-Jan-2020].

2018:-
1. Z. Doshi, S. Nadkarni, R. Agrawal and N. Shah, "AgroConsultant: Intelligent Crop
Recommendation System Using Machine Learning Algorithms," 2018 Fourth
International Conference on Computing Communication Control and Automation
(ICCUBEA), Pune, India, 2018, pp. 1-6. doi: 10.1109/ICCUBEA.2018.8697349.

2. K. Charvat et al., “Advanced Visualisation of Big Data for Agriculture as Part of

Databio Development,” in IGARSS 2018 - 2018 IEEE International Geoscience and
Remote Sensing Symposium, 2018, pp. 415–418, doi: 10.1109/IGARSS.2018.8517556.

2017:-
1. T. Edwin, “Onion, tomato price spike: season not the only reason”, Nov. 13, 2017. [Online].
Available: https://www.thehindubusinessline.com/economy/agri-business/onion-tomato-price-
spike-season-not-the-only-reason/article9957255.ece#. [Accessed Feb. 22, 2020].
2. Dhivya B H, Manjula R, Siva Bharathi S, Madhumathi R. A Survey on Crop Yield
Prediction based on Agricultural Data, International Journal of Innovative Research in
Science, Engineering and Technology. 2017; 6(3).

3.Jharna Majumdar, Sneha Naraseeyappa, Shilpa Ankalaki. Analysis of agriculture data using
datamining techniques: application of big data. Journal of Big data. 2017.

4. Swarupa Rani. The Impact of Data Analytics in Crop Management based on Weather
Conditions. International Journal of Engineering Technology Science and Research. 2017;
4(5):299-308.

5. DakshayiniPatil, Dr. M .S Shirdhonkar, “Rice Crop Yield Prediction using Data Mining
Techniques: An Overview”, International Journal of Advanced Research in Computer Science
and Software Engineering, Volume 7, Issue 5, ISSN: 2277 128X,2017.

6.Mukesh kumar , Prof.Mayura Nagar, "Big Data analytics in agriculture and distribution
channel," in IEEE International Conference on Computing Methodologies and
Communication, 2017.

7.Francesc X.Prenafeta-Boldu, Andreas Kamilaris, Andreas Kartakoullis, "A review on the

practice of big data analysis in agriculture," Computers and Electronics in Agriculture, vol.
143, no. C, p. 14, January 2017.
8. D.P.Acharjya, Kauser AhmedP, "A Survey on Big Data Analytics: Challenges, Open
Research Issues and Tools," International Journal of Advanced Computer Science and
Applications (IJACSA), vol. 7, no. 2, p. 9, 2017.

9.Van-Quyet Nguyen, Giang-Truong Nguyen, Sinh-Ngoc Nguyen, Jintae Choi, Kyungbaek

Kim, "Big Data Analytics and Visualization Techniques: A Case Study from Agriculture
Domain," in KISM Fall Conference, 2017.

10. Ch. Chandra Sekhar, J. Uday Kumar, B. Kishor Kumar, Ch. Sekhar, "Big Data Analytics on
Indian Crop Planning to Increase Agricultural Production," Advanced Science and Technology
Letters, vol. 147, p. 7, November 2017.

11. S. Rajeswari, K. Suthendran, and K. Rajakumar, “A smart agricultural model by integrating

IoT, mobile and cloud-based big data analytics,” in 2017 International Conference on
Intelligent Computing and Control (I2C2), 2017, pp. 1–5, doi:10.1109/I2C2.2017.8321902.

2016:-
1. S. Pudumalar, E. Ramanujam, R. H. Rajashree, C. Kavya, T. Kiruthika and J. Nisha,
"Crop recommendation system for precision agriculture," 2016 Eighth International
Conference on Advanced Computing (ICoAC), Chennai, 2017, pp. 32-36. doi:
 10.1109/ICoAC.2017.7951740.

2. T.R. Lekhaa, “Efficient Crop Yield and Pesticide Prediction for Improving Agricultural
Economy using Data Mining Techniques”, International Journal of Modern Trends in
Engineering and Science (IJMTES), 2016, Volume 03, Issue 10.

3. Majumdar J, Ankalaki S. Comparison of clustering algorithms using quality metrics

with invariant features extracted from plant leaves. International Conference on
Computational Science and Engineering. 2016.
4. 2015:-

1. R. Kumar, M. P. Singh, P. Kumar and J. P. Singh, "Crop Selection Method to maximize crop
yield rate using machine learning technique," 2015 International Conference on Smart
Technologies and Management for Computing, Communication, Controls, Energy and
Materials (ICSTM), Chennai, 2015, pp. 138-145. doi: 10.1109/ICSTM.2015.72254.