IAES International Journal of Artificial Intelligence (IJ-AI)

Vol. 11, No. 2, June 2022, pp. 753~763

ISSN: 2252-8938, DOI: 10.11591/ijai.v11.i2.pp753-763  753

A comprehensive review on machine learning in agriculture

domain

Kavita Jhajharia, Pratistha Mathur

Department of Information Technology, Faculty of Engineering, Manipal University Jaipur, Jaipur, India

Article Info ABSTRACT

Article history: Agriculture is an essential part of sustaining human life. Population growth,
climate change, resource competition are the key issues that increase food
Received Sep 6, 2021 security and to handle such complex problems in agriculture production,
Revised Feb 14, 2022 intelligent or smart farming extends the incorporation of technology into
Accepted Mar 3, 2022 traditional agriculture notion. Machine learning is a vitally used technology
in agriculture to protect food security and sustainability. Crop yield
production, water preservation, soil health and plant diseases can be
Keywords: addressed by machine learning. This paper has presented a compendious
review of research papers that deployed machine learning in the agriculture
Agriculture domain. The observed sub-categories of the agriculture domain are crop
Artificial neural network yield prediction, soil management, pest management, weed management,
Food security and crop disease. The outcomes represent that machine learning provides
Machine learning better accuracy concerning classification or regression. Machine learning
Support vector machine emerged with the internet of things, drones, robots, automated machinery,
and satellite imagery motivates researchers for smart farming and food
security.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Kavita Jhajharia
Department of Information Technology, Faculty of Engineering, Manipal University Jaipur
Dehmi Kalan, Near GVK Toll Plaza, Jaipur-Ajmer Expressway, Jaipur, Rajasthan 303007, India
Email: [email protected]

1. INTRODUCTION
Agriculture is a basic need for humankind to subsist. Continuous increment in population strains to
feed the ever-growing population. Resources and food production management is required to cater for the
augmented population. Agriculture production relies on many factors, such as soil type and quality, irrigation
management, weather, and water. Agriculture is a basic need for humankind to subsist. Continuous increment
in population strains to feed the ever-growing population. Resources and food production management is
required to cater for the augmented population. Farming has become more intensified to maximize crop
yields. To produce the sufficient amount of food, smart agriculture is required. Satellite data makes
agriculture more accurate and predictive. Smart farming has evolved widely in the last few years to fulfil the
food need.
Machine learning (ML) in consort with data analysis generates possibilities to understand and
reconnoitre the field of agriculture more effectually. According to Tom Michael, ML is a set of computer
instructions that learns from previous experience, concerning the task, and on the basis of previous
experience and task, performance is measured and which improves with experience and task [1]. Samuel
defines ML as a scientific domain of study which provides machines with the ability to learn without being
specifically programmed [2]. With time, machine learning is being widely applied in many fields, including
bioinformatics [3], anatomy [4], cheminformatics [5], economics [6], robot locomotion [7], speech

Journal homepage: http://ijai.iaescore.com

754  ISSN: 2252-8938

recognition [8], information retrieval [9], and neuroscience [10]. In this research paper, machine learning
algorithm in agriculture domain is deliberated [11].
The organization of the paper is: machine approach section has the description of machine learning
methods, techniques, and algorithms, the literature review section contains the review of the identified areas
of agriculture that have used machine learning, and discussion and conclusion section encloses the final
findings, conclusion and discussion of the paper along with the advantages of application of machine learning
in agriculture domain. ML is a process where the system or machine learns from experience and can improve
performance. Statistical and mathematical models can measure improved performance. Set of examples can
also be dictated as ML model or algorithms are trained using data sets. After the accomplishment of training,
the trained model is used to identify, predict or classify new input data. Figure 1 illustrates the ML approach.
ML algorithms explained below are not limited to the methods applied in papers used for this review process.

Figure 1. Machine learning approach

2. LITERATURE REVIEW
2.1. Research method
A systematic review methodology has been followed for the review conduction used in this research
paper. The review process includes review planning, search string, and search criteria for Machine learning in
agriculture. After completing the search, the paper selection is made based on inclusion and exclusion
criteria. This section contains information about how the review is accomplished.

2.2. Planning of review

Machine learning has evolved in agriculture rapidly in past years. However, despite numerous
research studies, the potential results for every field have not been identified yet. This review aims to provide
an outline of the machine learning technology in the agriculture domain and in-depth investigation. The work
analyses various sub-categories of the agriculture domain, techniques applied, observed features, and dataset
resources used in the research.

2.3. Search string

To conduct the search string, some keywords are identified as agriculture machine learning, ML
techniques agriculture, crop yield prediction machine learning, pest machine learning, crop disease machine
learning, soil machine learning, and weed machine learning, with the main emphasis on keywords machine
learning and agriculture. The authors performed an in-depth search to ensure the comprehensiveness of the
study. A few known papers may not have been considered because of title mismatch with the identified
keywords. Figure 2 represents the chosen search strings.

Int J Artif Intell, Vol. 11, No. 2, June 2022: 753-763

Int J Artif Intell ISSN: 2252-8938  755

2.4. Selection criteria

The literature review follows pre-specified selection criteria for including and excluding the papers
in the study. The inclusion criteria include the paper which matches the search string, and exclusion criteria
excluded the papers by title and domain mismatch, abstract and text irrelevance. Figure 3 illustrates the paper
inclusion and exclusion.

Figure 2. Search string

Figure 3. Inclusion and exclusion criteria

2.5. Review conduction

Machine learning is a game-changing technology and widely used in diversified fields. Machine
learning has been applied in the agricultural domain throughout the crop cycle. It starts with soil management
and ends with taking decisions about the crop's ripeness by the robot. In this review, articles have been
classified into the following categories: crop yield prediction, soil management, pest management, weed
management, and crop disease. The papers were searched using particular keywords for every selected
domain of agriculture. Agriculture has many sub-areas, and all cannot be included in the review; considering
this constraint, some areas are excluded. General abbreviations used in the paper are compiled in Table 1.

2.6. Categorical literature review

2.6.1. Crop yield prediction
In agriculture, crop yield, also known as agriculture output, is an essential component to complete
the growing population's need. Agriculture crop yield or productivity depends on many factors, such as
weather conditions, soil conditions, water, temperature, and rainfall. Therefore, ML can match the demand
and supply of food without affecting the environment or natural resources.

2.6.2. Soil management

Machine learning implementation has been used to predict and identify based on soil characteristics
such as valuation of soil moisture, condition, and temperature. A better prediction of soil condition can help
to improve soil management. ML technologies can achieve a more accurate estimation of soil with less time
and cost.

A comprehensive review machine learning in agriculture domain (Kavita Jhajharia)

756  ISSN: 2252-8938

Table 1. General abbreviations

Abbreviation Definition Abbreviation Definition
AMSR-E Advanced Microwave Scanning Radiometer on the Earth MODIS Moderate Resolution Imaging
Observing System Spectroradiometer
ANN Artificial Neural Network MPE Mean percent error
AI Artificial Intelligence NB Naïve Bayes
CNN Convolutional Neural Network NDVI Normalized difference vegetation index
CP-ANN Counter Propagation Artificial Neural Network NN Neural Network
DL Deep Learning PCA Principal Component Analysis
DT Decision Tree PLSDA Partial Least Squares Discriminant
Analysis
EL Ensemble Learning PMNN Perceptron Multilayer neural network
ELM Extreme Learning Machine RBF-NN Radial Basis Function Neural Network
EM Expectation Maximisation RE Relative Error
ERT Extremely randomized tree RF Random Forest
LR Logistic Regression RMSE Root mean square error
LS-SVM Least Squares Support Vector Machines SOM Self-Organizing Map
LSTM Long short-term memory SVM Support Vector Machine
ML Machine Learning SVR Support Vector Regression
MLR Multiple Linear Regression

2.6.3. Pest management

Pest damages the crops and reduces production, which can rigorously affect the food supply and
demand chain. Reduction of the crop damage and increment of the crop production compels the farmers to
use chemicals to control and protect the field from pests. Even though utilization of chemicals is harmful to
the environment, animals and human's health, ML algorithms can provide an efficient solution for pest
management.

2.6.4. Weed management

Weed in farming is the most undesirable plant that rivals the yield. It makes harvesting difficult and
includes impurity and moisture to crop. The negative effects of weeds on yields incorporate challenge to
sunlight, water, space, complex harvesting, and devaluation of crop quality. ML can detect weed on the crop.
Many articles have been presented here to detect and discriminate weed from the crop.

2.6.5. Crop disease

The rapidly increasing world population puts much pressure on agriculture resources. Crop
Production is the essential component to maintain the population need as well as the economic system. Crop
diseases are the primary source of plant damage, which affects crop production. Due to distressed climate and
environmental situations, a manifestation of plant illnesses is at the upward thrust. There are numerous crop
diseases and various symptoms containing spots/smudge appearing on plant leaves [12]. ML techniques
accommodated to detect the disease in the plant at an early stage. The Table 2 shown in appendix represents
the comparison of above-mentioned categories.

3. DISCUSSION
The review's primary focus is to brief the significant benefits of ML in the agriculture domain and
possible research areas. The review analyses the existing machine learning tools and techniques deployed in
the agriculture domain, including crop prediction, soil management, pest management, weed management
and crop disease. Many international journals cover the advances in the development and applications of
hardware, software, and related technologies for solving issues in the agriculture domain. The total number of
research articles reviewed is 38. The review includes 3 conference and 35 journal articles, as shown in
Figure 4. The presented articles here are from 2005 to till present, shown in Figure 5. The year-wise
distribution of reviewed papers is demonstrated in Figure 5. The result clearly shows that there is significant
work done in the last 3 to 4 years in agriculture using machine learning.
Analysis of the articles indicates that mainly nine ML algorithms are examined/adopted in the
survey, shown in Figure 6. In crop prediction, Nine ML algorithms are deployed; further analysis of the
surveyed articles indicates that ANN is the most popular algorithm applied in the field of crop prediction. In
soil management, five ML algorithms are deployed where SVM and regression are mainly used. In the pest
management category, five ML algorithms are deployed where SVM is majorly used. In Weed management,
five ML algorithms are implemented and, SVM is most often used. In last, crop disease, four ML algorithms
are implemented and, SVM is majorly used. Thus, the majority of work is done using ANN and SVM can be
concluded from the reviewed literature.

Int J Artif Intell, Vol. 11, No. 2, June 2022: 753-763

Int J Artif Intell ISSN: 2252-8938  757

8%

Journal Papers
92%
Conference Papers

Figure 4. Categorization of papers

10
8
Number of Papers

8
6
6 5 5
4
4 3
2 1 1 1 1 1 1 1
0
2005 2006 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Publication year

Figure 5. Number of papers published per years

5
Number of Papers

0
Crop Prediction Soil Pest Weed Crop Disease
Management Management Management

Agriculture Categories
Regression Decision Tree
Clustering Bayesian Algorithm
ANN DL
Ensemble Algorithm SVM
Instance Based Algorithm

Figure 6. Utilization of ML algorithms in different categories

The analysis of figures indicates that SVM is majorly implemented because of its sequential
approach, which incorporates several features to make a decision/ features into classes. SVM uses a kernel
function to differentiate the nonlinear and separable data and generates a mapping relationship between the
input vector and high-dimensional space vector through a hyperplane. SVM is preferred because of its sparse
representation and absence of local minima. Machine learning has a significant impact on application areas of
A comprehensive review machine learning in agriculture domain (Kavita Jhajharia)
758  ISSN: 2252-8938

the agriculture domain. Results produced by ML are promising. Particularly DL is getting more acceptance
because of its automatic feature extraction method in the agriculture sector, which can ease the process and
support the stakeholders of the agriculture domain. DL architectures/algorithms are also vastly implemented
in crop disease, weed management and crop prediction domains.

4. CONCLUSION
ML-based techniques have attracted much attention from researchers to improve the productivity in
agriculture domain. This review summarises the implementation of the ML algorithm in the agriculture
domain in the past few years. Though many algorithms are deployed, SVM and neural networks are the key
techniques to be better and precise. However, the researcher can explore new techniques, new domain, and
the inclusion of raw data to get more accurate results in the future. Deep learning is getting attention in the
past 3-4 years. The review covers five major domains; however, further study is required to explore the other
research areas of agriculture: rain management, weather Management, climate management, livestock
production, and animal welfare.

APPENDIX

Table 2. Comparison among multiple agriculture domains (continue)

Reference Agriculture Observed Functionality Applied Data Sources Results
No. Domain Features Algorithms
[13] Crop Seven-band Remote sensing LSTM, Moderate resolution Trained and tested
prediction reflectance data used to train regression imaging spectroradiometer the model on
imagery the model to satellite imagery soybean data of
predict the crop Argentina and
yield of one predicts fine for
region. Then, brazil.
another region Pre-trained model
prediction was
performed using
transfer learning
[14] Crop Soil Estimates crop SVM, ERT, National Agricultural DL produced the
prediction moisture yield and present RF, DL Statistical Service and highest accuracy
comparison among United States of among all
many machine Department of Agriculture,
learning National Aeronautics and
techniques Space Administration,
European Space Agency,
Climate Change Initiative
and PRISM Climate Group
[15] Crop Multiple Detects each X-Means, 154 images were collected Recall value: 0.80
prediction features integral tomato DT by conventional RGB Precision: 0.88
color, shape, fruit which digital camera at Tsukuba Recall of young
texture and incorporates Plant Factory of the fruit: 0.78
size mature, immature, Institute of Vegetable and
and young fruits Tea Science, Ibaraki, Japan
on a tomato plant
[16] Crop Multilayer Predicts wheat CP-ANN, Duck End Farm Field, Accuracy:
prediction soil yield for three XY-Fusion, Wilstead, Bedfordshire, U. Supervised kohonen
parameters isofrequency Supervised K. network: 81.65%
classes, namely Kohonen CP-ANN: 78.3%
high, medium and Network XY-Fusion: 80.92%
low
[17] Crop Geometrical Detects tomatoes K-Means, RGB images of Spatial K-means
prediction features from RGB images SOM, EM resolution acquired from Precision: 0.723
unmanned aerial vehicles Recall: 0.593
F-Measure: 0.652
SOM
Precision: 0.730
Recall: 0.686
F-Measure: 0.707
EM
Precision: 0.919
Recall: 0.606
F-Measure: 0.730

Int J Artif Intell, Vol. 11, No. 2, June 2022: 753-763

Int J Artif Intell ISSN: 2252-8938  759

Table 2. Comparison among multiple agriculture domains (continue)

Reference Agriculture Observed Functionality Applied Data Sources Results
No. Domain Features Algorithms
[18] Crop Soil SBOCM used to SVM Chinese Middle-season rice
prediction properties predict different Academy of Tillering stage:
stages and yield of Sciences RE(%)=22.1
rice Heading stage:
RE(%)=17.1
Milk stage:
RE(%)=19.2

Early rice Tillering stage:

RE(%)=20.5 Heading stage:
RE(%)=15.8
Milk stage:
RE(%)=8.5

Late rice: Tillering stage:

RE(%)=21.0
Heading stage:
RE(%)=16.5
Milk stage:
RE(%)=11.1
[19] Crop Irrigation Crop yield MLR, M5- Irrigation M5-Prime predicted with the
prediction water, prediction Prime module of best accuracy, followed by
rainfall, performed for two Regression Santa Rosa KNN, SVR and MLR.
temperature consecutive years Trees, [Agricultural
PMNN, Production
SVR, K-NN Data and
Weather
information
Data]
[20] Crop Vegetation Determines the ANN Emile A. Lods RMSE (kg/ha)= 19.7
prediction indices potential of Agronomy
hyperspectral data Research
and ANNs Centre data
obtained by
Compact
Airborne
Spectrographic
Image
[21] Soil N/A Predict soil texture SVM, ANN Tuscany, RMSE:
management and stoniness Central Italy SVM
based on γ- Sand: 7.0
spectroscopy Clay: 5.9
Stoniness:0.10
ANN
Sand:7.9
Clay:6.3
Stoniness:0.11
[22] Soil N/A Crop yield Stepwise Lower seyhan MPE:
management prediction based linear plane, berdan, Wheat: 7.9%
on soil salinity regression seyhan, and Corn: 8.8%
ceyhan rivers Cotton:6.3%
Crop Yield loss:
Corn: 55%
wheat: 28%
Cotton: 15%
[23] Soil N/A AMSR-E data is RF Global change Coefficient correlation (r)
management consistently used master South korea:0.71
to observe patterns directory and Australis: 0.84
of Global soil rural RMSE:
moisture development South Korea:0.049
administration Australia: 0.05
[24] Soil N/A Uses near-infrared LS-SVM, Top soil layer RMSE of prediction
management and visible bands Cubist from Premslin, LS-SVM:
to predict soil Germany. Moisture content: 0.457%
nitrogen, organic Organic carbon: 0.062%
carbon, and
moisture
[25] Soil N/A Predicts soil SVM Chi-Chi, Performance: 77.65%
management liquefaction Taiwan
susceptibility earthquake.

A comprehensive review machine learning in agriculture domain (Kavita Jhajharia)

760  ISSN: 2252-8938

Table 2. Comparison among multiple agriculture domains (continue)

Reference Agriculture Observed Functionality Applied Data Sources Results
No. Domain Features Algorithms
[26] Soil N/A Implemented digital Cubist, RF Borujen region, soil organic carbon:
management soil mapping Chaharmahal-Va- RMSE: 0.33(RF)
techniques to estimate Bakhtiari Province, calcium carbonate
the spatial distribution central Iran equivalent
of numerous soil RMSE: 9.52(Cubist)
properties Clay:
RMSE: 7.86(RF)
[27] Soil N/A Defines and assesses CNN LUCAS Soil RMSE
management the efficiency of database Organic carbon: 10.5%
transfer learning to Cation exchange
localize capacity: 11.8%
Clay content: 12.0%
pH: 11.5%
[28] Pest Color, Automated rice pest SVM Live images with Accuracy 97.5%
management Shape, identification system cameras
Texture
[29] Pest Area, Detect individual pest ANN Sugar beet field in R=0.89
management Perimeter, among other species Shiraz, Iran
sphericity,
Eccentricity
[30] Pest Curve Diagnosis of plant pest SVM Lancaster Tomato (mildew)
management response using Electronic nose. University, UK Linear: 95%
and slope Polynomial: 94%
RBF: 96%
Cucumber (wounded)
Linear: 77%
Polynomial: 82%
RBF: 87%
Cucumber (spider mite)
Linear: 94%
Polynomial: 88%
RBF: 91%
Pepper (wounded)
Linear: 67%
Polynomial:71%
RBF: 92%
[31] Pest 58 attributes Develop a method to AdaBoost, Zespri International Precision:
management forecast the result of NB Ltd AdaBoost: 98%
pest monitoring. Naïve Bayes: 95%
[32] Pest N/A Detects and classifies DL 88,670 images Mean average Precision:
management multi-class pests. 75.46%
[33] Pest color Automatically detects SVM Tarbiat Modares MPE of less than 2.25%
management indexes thrips and their University,
were: Hue, position. Islamic Republic of
Saturation Iran, Tehran
and
Intensify
[34] Weed Color, Pynovisao software CNN Images captured by CNN:
management shape, developed and used to unmanned aerial Precision 0.991
texture and detect weed in crop vehicle. Sensitivity 0.991
image image and classified
orientation using CNN.
[35] Weed Nitrogen Weed classification SVM 72-waveband Effect of nitrogen and
management application performed w.r.t. compact airborne weed combined: 69.2%
rate: 60,120 nitrogen application spectrographic Effect of nitrogen:80.8
and 250 kg rate imager (CASI), Effect of weed: 85.8
N/ha range: 408.73 to
947.07 nm
[36] Weed Color and Weed discrimination DT Rice and weed Precision: 0.982
management texture for different growing images from the Recall: 0.977
states of rice internet of
1125*1500
[37] Weed Spectral Recognizes weed SOM, Hyperspectral Mixture of Gaussian-
management species based on Mixture of images using HSI. 31%-98%
hyperspectral sensing. Gaussian SOM- 53%-94%
[38] Weed Color, Weed and crop were SVM OLYMPUS FE4000 Accuracy- 97%
managemen moment classified using digital point-and-shoot
invariant, images. digital camera
size

Int J Artif Intell, Vol. 11, No. 2, June 2022: 753-763

Int J Artif Intell ISSN: 2252-8938  761

Table 2. Comparison among multiple agriculture domains

Reference Agriculture Observed Functionality Applied Data Sources Results
No. Domain Features Algorithms
[39] Weed Shape, Fourier Weed detection SVM, ANN 960×1280 pixels, Accuracy:
management descriptor, using shape features Shiraz university. ANN: 92.92%
moment SVM: 95.00%
invariant
[40] Weed RGB-NIR Detect sugar beet CNN UAVs equipped Accuracy 95%
management imagery plant and weed- with vision
based on vision sensors
classification
[41] Weed Size, length, and Classification for SVM Red (580 nm) and Overall accuracy: 97.7%
management fourier small-grain weed infrared (>720
species concerning nm) spectrum
cirsium arvense and
galium aparine
[42] Crop disease Hyperspectral Detecting sclerotinia PLSDA, farm of Zhejiang Sample set 1:
imaging with 2.8 sclerotiorum on RBF-NN, University Average spectra:
mm spectral oilseed rape stems SVM, and PLSDA: 100
resolution, pixel ELM RBFNN: 97.50
size is 6.45×6.45 ELM: 100
µm SVM: 92.50
Pixel-wise Spectra:
PLSDA: 94.80
RBFNN: 98.80
ELM: 99.40
SVM: 99.00

Sample set 2:
Average spectra:
PLSDA: 92.50
RBFNN: 87.50
ELM: 97.50
SVM: 90.00
Pixel-wise Spectra:
PLSDA: 96.60
RBFNN: 98.70
ELM: 99.50
SVM: 99.30
[43] Crop disease Leaf, stem, and Detect real-time DL Images using a Mean average precision
fruits disease along with digital camera 83.06%
the class and from farms of the
location of the plant Korean peninsula
[44] Crop disease Spectral Detects and SVM Cercospora leaf Cercospora Leaf spot:
vegetation classifies plant spot, leaf rust and 89.69
indices diseases in sugar powdery mildew Sugar beet rust: 83.60
beet Powdery mildew: 92.46
[45] Crop disease Coloured, Detects plant disease CNN PlantVillage Overall accuracy-
greyscale and using images Public dataset 99.35%
segmented
[46] Crop disease 75 features by Healthy and KNN GAP Agricultural KNN:
wavelet fusarium diseased research Statistics of wavelet
decomposition pepper leaves were (GAPTEAM), coefficient: 99%
detected şanlıurfa, Turkey Wavelet Coefficient:
100%
[47] Crop disease Grayscale Detect and classify CNN Images captured Dataset split:
potato disease by by cameras 90%-train and 10%-test
visible symptoms provides accuracy -
0.9585
[48] Crop disease Shape, texture, Identification of SVM The University of Accuracy 93.1%
and grey level plant disease by Georgia, USA
visual symptoms
[49] Crop disease Leaf properties Classifies the CNN Plant village Accuracy 99.18%
disease based on
symptoms visible
[50] Crop disease Color, texture, Detects disease in ANN, ANN, SVR-RBF, RMSE:
gray level co- apple fruit SVR-rbf, and SVR-Poly ANN: 0.53
occurrence and SVR- SVR-Poly: 0.42
matrix, and Poly SVR-RBF: 0.2
wavelet
transform

A comprehensive review machine learning in agriculture domain (Kavita Jhajharia)

762  ISSN: 2252-8938

REFERENCES
[1] T. M. Mitchell, Machine Learning. New York: McGraw-Hill, 1997.
[2] A. L. Samuel, “Some studies in machine learning using the game of checkers. I,” in Computer Games I, vol. 3, no. 3, C. G. I and
D. N. L. Levy, Eds. New York, NY: Springer New York, 1988, pp. 335–365.
[3] I. Inza, B. Calvo, R. Armañanzas, E. Bengoetxea, P. Larrañaga, and J. A. Lozano, “Machine learning: an indispensable tool in
bioinformatics,” Methods in Molecular Biologyiology, vol. 593, pp. 25–48, 2010, doi: 10.1007/978-1-60327-194-3_2.
[4] X. Zhu, Y. Ge, T. Li, D. Thongphiew, F.-F. Yin, and Q. J. Wu, “A planning quality evaluation tool for prostate adaptive IMRT
based on machine learning,” Medical Physics, vol. 38, no. 2, pp. 719–726, Jan. 2011, doi: 10.1118/1.3539749.
[5] Y.-C. Lo, S. E. Rensi, W. Torng, and R. B. Altman, “Machine learning in chemoinformatics and drug discovery,” Drug Discovery
Today, vol. 23, no. 8, pp. 1538–1546, Aug. 2018, doi: 10.1016/j.drudis.2018.05.010.
[6] H. Park, N. Kim, and J. Lee, “Parametric models and non-parametric machine learning models for predicting option prices:
empirical comparison study over KOSPI 200 Index options,” Expert Systems with Applications, vol. 41, no. 11, pp. 5227–5237,
Sep. 2014, doi: 10.1016/j.eswa.2014.01.032.
[7] N. Kohl and P. Stone, “Policy gradient reinforcement learning for fast quadrupedal locomotion,” in IEEE International
Conference on Robotics and Automation, 2004., 2004, no. 3, pp. 2619–2624, doi: 10.1109/ROBOT.2004.1307456.
[8] X. Xu, J. Deng, E. Coutinho, C. Wu, L. Zhao, and B. W. Schuller, “Connecting subspace learning and extreme learning machine
in speech emotion recognition,” IEEE Transactions on Multimedia, vol. 21, no. 3, pp. 795–808, Mar. 2019, doi:
10.1109/TMM.2018.2865834.
[9] F. Sebastiani, “Machine learning in automated text categorization,” ACM Computing Surveys, vol. 34, no. 1, pp. 1–47, Mar. 2002,
doi: 10.1145/505282.505283.
[10] J. Richiardi, S. Achard, H. Bunke, and D. Van De Ville, “Machine learning with brain graphs: predictive modeling approaches for
functional imaging in systems neuroscience,” IEEE Signal Processing Magazine, vol. 30, no. 3, pp. 58–70, May 2013, doi:
10.1109/MSP.2012.2233865.
[11] K. Liakos, P. Busato, D. Moshou, S. Pearson, and D. Bochtis, “Machine learning in agriculture: a review,” Sensors, vol. 18, no. 8,
Aug. 2018, doi: 10.3390/s18082674.
[12] P. Chandana et al., “An effective identification of crop diseases using faster region based convolutional neural network and expert
systems,” International Journal of Electrical and Computer Engineering (IJECE), vol. 10, no. 6, pp. 6531–6540, Dec. 2020, doi:
10.11591/ijece.v10i6.pp6531-6540.
[13] A. X. Wang, C. Tran, N. Desai, D. Lobell, and S. Ermon, “Deep transfer learning for crop yield prediction with remote sensing
data,” in Proceedings of the 1st ACM SIGCAS Conference on Computing and Sustainable Societies, Jun. 2018, pp. 1–5, doi:
10.1145/3209811.3212707.
[14] N. Kim and Y.-W. Lee, “Machine learning approaches to corn yield estimation using satellite images and climate data: a case of
iowa state,” Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography, vol. 34, no. 4, pp. 383–390,
Aug. 2016, doi: 10.7848/ksgpc.2016.34.4.383.
[15] K. Yamamoto, W. Guo, Y. Yoshioka, and S. Ninomiya, “On plant detection of intact tomato fruits using image analysis and
machine learning methods,” Sensors, vol. 14, no. 7, pp. 12191–12206, Jul. 2014, doi: 10.3390/s140712191.
[16] X. E. Pantazi, D. Moshou, T. Alexandridis, R. L. Whetton, and A. M. Mouazen, “Wheat yield prediction using machine learning
and advanced sensing techniques,” Computers and Electronics in Agriculture, vol. 121, pp. 57–65, Feb. 2016, doi:
10.1016/j.compag.2015.11.018.
[17] J. Senthilnath, A. Dokania, M. Kandukuri, R. K.N., G. Anand, and S. N. Omkar, “Detection of tomatoes using spectral-spatial
methods in remotely sensed RGB images captured by UAV,” Biosystems Engineering, vol. 146, pp. 16–32, Jun. 2016, doi:
10.1016/j.biosystemseng.2015.12.003.
[18] Y. Su, H. Xu, and L. Yan, “Support vector machine-based open crop model (SBOCM): case of rice production in China,” Saudi
Journal of Biological Sciences, vol. 24, no. 3, pp. 537–547, Mar. 2017, doi: 10.1016/j.sjbs.2017.01.024.
[19] A. Gonzalez-Sanchez, J. Frausto-Solis, and W. Ojeda-Bustamante, “Predictive ability of machine learning methods for massive
crop yield prediction,” Spanish Journal of Agricultural Research, vol. 12, no. 2, pp. 313–328, Apr. 2014, doi:
10.5424/sjar/2014122-4439.
[20] Y. Uno et al., “Artificial neural networks to predict corn yield from compact airborne spectrographic imager data,” Computers
and Electronics in Agriculture, vol. 47, no. 2, pp. 149–161, May 2005, doi: 10.1016/j.compag.2004.11.014.
[21] S. Priori, N. Bianconi, and E. A. C. Costantini, “Can γ-radiometrics predict soil textural data and stoniness in different parent
materials? a comparison of two machine-learning methods,” Geoderma, vol. 226–227, no. 1, pp. 354–364, Aug. 2014, doi:
10.1016/j.geoderma.2014.03.012.
[22] O. Satir and S. Berberoglu, “Crop yield prediction under soil salinity using satellite derived vegetation indices,” Field Crops
Research, vol. 192, pp. 134–143, Jun. 2016, doi: 10.1016/j.fcr.2016.04.028.
[23] J. Im, S. Park, J. Rhee, J. Baik, and M. Choi, “Downscaling of AMSR-E soil moisture with MODIS products using machine
learning approaches,” Environmental Earth Sciences, vol. 75, no. 15, Aug. 2016, doi: 10.1007/s12665-016-5917-6.
[24] A. Morellos et al., “Machine learning based prediction of soil total nitrogen, organic carbon and moisture content by using VIS-
NIR spectroscopy,” Biosystems Engineering, vol. 152, pp. 104–116, Dec. 2016, doi: 10.1016/j.biosystemseng.2016.04.018.
[25] P. Samui and T. G. Sitharam, “Machine learning modelling for predicting soil liquefaction susceptibility,” Natural Hazards and
Earth System Sciences, vol. 11, no. 1, pp. 1–9, Jan. 2011, doi: 10.5194/nhess-11-1-2011.
[26] M. Zeraatpisheh, S. Ayoubi, A. Jafari, S. Tajik, and P. Finke, “Digital mapping of soil properties using multiple machine learning
in a semi-arid region, central Iran,” Geoderma, vol. 338, pp. 445–452, Mar. 2019, doi: 10.1016/j.geoderma.2018.09.006.
[27] J. Padarian, B. Minasny, and A. B. McBratney, “Transfer learning to localise a continental soil vis-NIR calibration model,”
Geoderma, vol. 340, pp. 279–288, Apr. 2019, doi: 10.1016/j.geoderma.2019.01.009.
[28] Q. Yao et al., “An insect imaging system to automate rice light-trap pest identification,” Journal of Integrative Agriculture, vol.
11, no. 6, pp. 978–985, Jun. 2012, doi: 10.1016/S2095-3119(12)60089-6.
[29] K. A. Vakilian and J. Massah, “Performance evaluation of a machine vision system for insect pests identification of field crops
using artificial neural networks,” Archives Of Phytopathology And Plant Protection, vol. 46, no. 11, pp. 1262–1269, Jul. 2013,
doi: 10.1080/03235408.2013.763620.
[30] R. Ghaffari et al., “Plant pest and disease diagnosis using electronic nose and support vector machine approach,” Journal of Plant
Diseases and Protection, vol. 119, no. 5–6, pp. 200–207, 2012, doi: 10.1007/BF03356442.
[31] M. G. Hill, P. G. Connolly, P. Reutemann, and D. Fletcher, “The use of data mining to assist crop protection decisions on
kiwifruit in New Zealand,” Computers and Electronics in Agriculture, vol. 108, pp. 250–257, Oct. 2014, doi:

Int J Artif Intell, Vol. 11, No. 2, June 2022: 753-763

Int J Artif Intell ISSN: 2252-8938  763

10.1016/j.compag.2014.08.011.
[32] L. Liu et al., “PestNet: an end-to-end deep learning approach for large-scale multi-class pest detection and classification,” IEEE
Access, vol. 7, pp. 45301–45312, 2019, doi: 10.1109/ACCESS.2019.2909522.
[33] M. A. Ebrahimi, M. H. Khoshtaghaza, S. Minaei, and B. Jamshidi, “Vision-based pest detection based on SVM classification
method,” Computers and Electronics in Agriculture, vol. 137, pp. 52–58, May 2017, doi: 10.1016/j.compag.2017.03.016.
[34] A. dos S. Ferreira, D. M. Freitas, G. G. da Silva, H. Pistori, and M. T. Folhes, “Weed detection in soybean crops using
ConvNets,” Computers and Electronics in Agriculture, vol. 143, pp. 314–324, Dec. 2017, doi: 10.1016/j.compag.2017.10.027.
[35] Y. Karimi, S. O. Prasher, R. M. Patel, and S. H. Kim, “Application of support vector machine technology for weed and nitrogen
stress detection in corn,” Computers and Electronics in Agriculture, vol. 51, no. 1–2, pp. 99–109, Apr. 2006, doi:
10.1016/j.compag.2005.12.001.
[36] B. Cheng and E. T. Matson, “A feature-based machine learning agent for automatic rice and weed discrimination,” in Lecture
Notes in Artificial Intelligence (Subseries of Lecture Notes in Computer Science), vol. 9119, L. Rutkowski, M. Korytkowski, R.
Scherer, R. Tadeusiewicz, L. A. Zadeh, and J. M. Zurada, Eds. Cham: Springer International Publishing, 2015, pp. 517–527.
[37] X.-E. Pantazi, D. Moshou, and C. Bravo, “Active learning system for weed species recognition based on hyperspectral sensing,”
Biosystems Engineering, vol. 146, pp. 193–202, Jun. 2016, doi: 10.1016/j.biosystemseng.2016.01.014.
[38] F. Ahmed, H. A. Al-Mamun, A. S. M. H. Bari, E. Hossain, and P. Kwan, “Classification of crops and weeds from digital images:
a support vector machine approach,” Crop Protection, vol. 40, pp. 98–104, Oct. 2012, doi: 10.1016/j.cropro.2012.04.024.
[39] A. Bakhshipour and A. Jafari, “Evaluation of support vector machine and artificial neural networks in weed detection using shape
features,” Computers and Electronics in Agriculture, vol. 145, pp. 153–160, Feb. 2018, doi: 10.1016/j.compag.2017.12.032.
[40] A. Milioto, P. Lottes, and C. Stachniss, “Real-time blob-wise sugar beets vs weeds classification for monitoring fields using
convolutional neural networks,” ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol.
IV-2/W3, no. 2W3, pp. 41–48, Aug. 2017, doi: 10.5194/isprs-annals-IV-2-W3-41-2017.
[41] T. Rumpf, C. Römer, M. Weis, M. Sökefeld, R. Gerhards, and L. Plümer, “Sequential support vector machine classification for
small-grain weed species discrimination with special regard to cirsium arvense and galium aparine,” Computers and Electronics
in Agriculture, vol. 80, pp. 89–96, Jan. 2012, doi: 10.1016/j.compag.2011.10.018.
[42] W. Kong, C. Zhang, W. Huang, F. Liu, and Y. He, “Application of hyperspectral imaging to detect sclerotinia sclerotiorum on
oilseed rape stems,” Sensors, vol. 18, no. 2, Jan. 2018, doi: 10.3390/s18010123.
[43] A. Fuentes, S. Yoon, S. Kim, and D. Park, “A robust deep-learning-based detector for real-time tomato plant diseases and pests
recognition,” Sensors, vol. 17, no. 9, Sep. 2017, doi: 10.3390/s17092022.
[44] T. Rumpf, A.-K. Mahlein, U. Steiner, E.-C. Oerke, H.-W. Dehne, and L. Plümer, “Early detection and classification of plant
diseases with support vector machines based on hyperspectral reflectance,” Computers and Electronics in Agriculture, vol. 74, no.
1, pp. 91–99, Oct. 2010, doi: 10.1016/j.compag.2010.06.009.
[45] S. P. Mohanty, D. P. Hughes, and M. Salathé, “Using deep learning for image-based plant disease detection,” Frontiers in Plant
Science, vol. 7, no. 6, pp. 1083–1087, Sep. 2016, doi: 10.3389/fpls.2016.01419.
[46] K. Karadağ, M. E. Tenekeci, R. Taşaltın, and A. Bilgili, “Detection of pepper fusarium disease using machine learning algorithms
based on spectral reflectance,” Sustainable Computing: Informatics and Systems, vol. 28, Dec. 2020, doi:
10.1016/j.suscom.2019.01.001.
[47] D. Oppenheim, G. Shani, O. Erlich, and L. Tsror, “Using deep learning for image-based potato tuber disease detection,”
Phytopathology, vol. 109, no. 6, pp. 1083–1087, Jun. 2019, doi: 10.1094/PHYTO-08-18-0288-R.
[48] A. Camargo and J. S. Smith, “Image pattern classification for the identification of disease causing agents in plants,” Computers
and Electronics in Agriculture, vol. 66, no. 2, pp. 121–125, May 2009, doi: 10.1016/j.compag.2009.01.003.
[49] M. Brahimi, K. Boukhalfa, and A. Moussaoui, “Deep learning for tomato diseases: classification and symptoms visualization,”
Applied Artificial Intelligence, vol. 31, no. 4, pp. 299–315, Apr. 2017, doi: 10.1080/08839514.2017.1315516.
[50] E. Omrani, B. Khoshnevisan, S. Shamshirband, H. Saboohi, N. B. Anuar, and M. H. N. M. Nasir, “Potential of radial basis
function-based support vector regression for apple disease detection,” Measurement, vol. 55, pp. 512–519, Sep. 2014, doi:
10.1016/j.measurement.2014.05.033.

BIOGRAPHIES OF AUTHORS

Kavita Jhajharia was Born in Jhunjhunu, Rajasthan, India, in 1992. She Received
her B.Tech. Degree from Rajasthan Technical University, India, in 2013 in Information
Technology, and the M.Tech Degree from SRM University, Sonepat, India, in 2016. She is
Assistant Professor in Manipal University Jaipur since 2016. She is member of ACM. Her
Research interest is VANET, Wireless networking, Machine Learning, Software Engineering
and IOT. She can be contacted at email: [email protected].

Pratistha Mathur is a Professor in the Department of Information Technology at

Manipal University Jaipur. She had an experience of more than 20 years. She has done her Ph.D.
in computer Science from Banasthali Vidyapith in 2012. She has done M.Tech. in computer
science in 1998 and secure the gold medal. Her Research areas are Digital Image Processing,
Soft Computing and Machine Learning. She is currently guiding many scholars at Ph.D. and
M.Tech. level. She has also been worked in the area of Indian language computing and
associated with many funded projects of MCIT, DOE Rajasthan, DST Rajasthan. She has
published more than 40 paper in international and national journals and conferences. She has
also attended more than 25 workshops and trainings. She can be contacted at email:
[email protected].

A comprehensive review machine learning in agriculture domain (Kavita Jhajharia)