Scribd
Upload
43 views8 pages

Crop Yield Prediction Using Deep Learning and Data

The project aims to develop a deep learning-based system for predicting crop yield using historical data, analyzing factors like weather patterns and soil conditions. It involves data collection, preprocessing, model development, evaluation, and visualization to assist farmers in making data-driven decisions. The project will utilize Python and various libraries for implementation and is structured over an eight-week timeline for completion.

Uploaded by

jgokulraj011
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
43 views8 pages

Crop Yield Prediction Using Deep Learning and Data

The project aims to develop a deep learning-based system for predicting crop yield using historical data, analyzing factors like weather patterns and soil conditions. It involves data collection, preprocessing, model development, evaluation, and visualization to assist farmers in making data-driven decisions. The project will utilize Python and various libraries for implementation and is structured over an eight-week timeline for completion.

Uploaded by

jgokulraj011
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd

CROP YIELD PREDICTION

USING DEEP LEARNING


AND DATA
VISUALIZATION
TEAM MEMBERS
ABISHEK M (2117230020005)
ARAVIND N (2117230020017)
BHUVANESH KUMAR M (2117230020030)
III YEAR CSE ‘A’ SEC

MENTOR: Dr. C. GETHARA GOWRI, AP/CSE


PROJECT OBJECTIVES
 TO DEVELOP A DEEP LEARNING-BASED SYSTEM THAT ACCURATELY PREDICTS
CROP YIELD USING HISTORICAL DATA.
 TO ANALYZE THE INFLUENCE OF WEATHER PATTERNS, SOIL CONDITIONS,
AND FARMING PRACTICES ON CROP PRODUCTION.
 TO ASSIST FARMERS AND STAKEHOLDERS WITH DATA-DRIVEN DECISIONS
FOR BETTER PLANNING AND SUSTAINABILITY.
 TO PRESENT THE RESULTS THROUGH INTERACTIVE AND INSIGHTFUL
VISUALIZATIONS FOR EASIER UNDERSTANDING AND IMPACT ANALYSIS.
PROPOSED METHODOLOGY / APPROACH
o DATA COLLECTION
 GATHER DATA ON RAINFALL, TEMPERATURE, SOIL TYPE, CROP TYPE, AND
HISTORICAL YIELD.
 SOURCES: KAGGLE, GOVERNMENT PORTALS (DATA.GOV.IN, FAO, ETC.)

o DATA PREPROCESSING
 CLEAN AND NORMALIZE DATA.
 HANDLE MISSING VALUES AND ENCODE CATEGORICAL VARIABLES.

o MODEL DEVELOPMENT:
 DESIGN A DEEP LEARNING MODEL (PREFERABLY LSTM OR ANN) FOR
PREDICTION.
 TRAIN ON HISTORICAL DATA WITH SUPERVISED LEARNING
PROPOSED METHODOLOGY / APPROACH
– CONTD.
o MODEL EVALUATION:
 EVALUATE PERFORMANCE USING METRICS LIKE RMSE, MAE, AND
ACCURACY.

o VISUALIZATION:
 USE VISUALIZATION LIBRARIES (E.G., MATPLOTLIB, PLOTLY) TO PRESENT
 RESULTS AND CORRELATIONS

o RESULT INTERPRETATION & DEPLOYMENT


 INTERPRET MODEL OUTPUTS AND OPTIONALLY DEPLOY VIA WEB
DASHBOARD OR GUI.
TECHNOLOGY / TOOLS TO BE USED
o LANGUAGES AND LIBRARIES:
 PYTHON: CORE PROGRAMMING LANGUAGE
 PANDAS, NUMPY: DATA MANIPULATION
 MATPLOTLIB, SEABORN, PLOTLY: DATA VISUALIZATION
 TENSORFLOW/KERAS: DEEP LEARNING MODEL CREATION
 SCIKIT-LEARN: ADDITIONAL PREPROCESSING AND METRICS
 OPENCV (OPTIONAL): IF INTEGRATING SATELLITE OR DRONE IMAGES IN FUTURE
PHASES

o PLATFORMS:
 GOOGLE COLAB / JUPYTER NOTEBOOK: DEVELOPMENT AND EXPERIMENTATION
 GITHUB: VERSION CONTROL
 KAGGLE / DATA.GOV.IN
INITIAL LITERATURE SURVEY
(REFERENCES)
• PAPER 1: “DEEP LEARNING FOR CROP YIELD PREDICTION: A REVIEW”
EMPHASIZES LSTM'S PERFORMANCE ON TIME-SERIES AGRI-DATA.

• PAPER 2: “MACHINE LEARNING MODELS FOR YIELD FORECASTING”


COMPARATIVE STUDY OF ML MODELS; DEEP LEARNING OUTPERFORMS
REGRESSION MODELS.

• PAPER 3: “DATA-DRIVEN AGRICULTURE: CHALLENGES AND


OPPORTUNITIES”
SHOWS BENEFITS OF VISUAL ANALYTICS IN AGRI-TECH.

• KAGGLE DATASETS:
CROP YIELD PREDICTION CHALLENGE INDIA AGRICULTURE CROP
WORK PLAN / TIMELINE
Week Tasks
Week 1 Finalize title, form team, assign roles
Week 2 Collect datasets from Kaggle, government
sources
Week 3 Preprocess and clean data
Week 4 Explore features, design deep learning model
Week 5 Train and validate the model
Week 6 Develop visualizations for insights
Week 7 Combine model + visual interface
Week 8 Final testing, documentation, and
review preparation
THANK YOU

You might also like