 Coastal regions in India are very frequently hit
byTropical Cyclones, which result in tremendous loss.
 Its intensity prediction has been a challenging task because of
drastic climatic changes over the past few years in the world.
 Intensity of Tropical Cyclone is highly influenced by ocean,
atmospheric and meteorological parameters which makes the
task difficult to define the mechanism of Tropical Cyclone
intensity prediction.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 1
 To develop hybrid learning models that can improve the accuracy of tropical
cyclone intensity predictions. By combining multiple machine learning or
statistical techniques(using both empirical and numerical modelling techniques)
 To develop models that can provide reliable and timely predictions of cyclone
intensification well in advance. By improving the lead time and warning systems,
researchers aim to help communities and authorities make better-informed
decisions and take appropriate actions to mitigate the impact of tropical
cyclones.
 To quantify and address uncertainties associated with predictions. The objective
is to provide not only point forecasts but also probabilistic estimates, which can
help decision-makers understand the range of potential outcomes and associated
risks.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 2
 Here, a hybrid deep learning model is built using historical observations
collected from various sources to perform a data-driven prediction
ofTropical Cyclone’s intensity using regression model.
 This hybrid model utilizes Convolutional Neural Network architectures for
feature extraction and machine learning models for regression.
 Application and Real-Time Implementation: The study may explore the
practical application of the developed hybrid learning models in real-time
forecasting and operational settings.
 This could involve implementing the models within existing weather
forecasting systems or assessing their feasibility for integration into decision
support systems used by meteorological agencies or disaster management
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 3
 Varalakshmi.P et al. investigated Tropical cyclone
intensity prediction based on hybrid learning techniques.
 Xin Wang et al. proposed Tropical cyclone intensity
change prediction based on surrounding environmental
conditions with Deep learning.
 Gao S et al. explained Improvements in typhoon intensity
change classification by incorporating an ocean coupling
potential intensity index into decision trees.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 4
 Li Y et al. explored Spectral spatial classification of hyper
spectral imagery with 3D convolutional neural network.
 Lee J ability to precisely extract Tropical cyclone intensity
estimation using multi-dimensional convolutional neural
networks from geostationary satellite data.
 Dvorak V et al. proposed Tropical cyclone intensity
analysis and forecasting from satellite imagery.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 5
 The existing literature provides a strong foundation for the proposed
study on tropical cyclone intensity prediction using hybrid learning
techniques.
 By building upon the previous research, this study aims to contribute
to the field by developing novel hybrid models, incorporating diverse
data sources, addressing uncertainty, and improving lead time and
warning systems.
 The findings from this work can have significant implications for
disaster management and community resilience in the face of tropical
cyclones.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 6
 Tropical cyclones pose significant threats to coastal
communities and require accurate intensity prediction for
effective disaster preparedness and response.
 This research proposal aims to explore the application of
hybrid learning techniques in improving tropical cyclone
intensity prediction.
 By combining multiple algorithms and integrating diverse
data sources, we aim to enhance forecast accuracy and
provide valuable insights into cyclone dynamics.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 7
 Design and develop hybrid learning models using a
combination of machine learning algorithms, statistical
methods, and physical models. Image processing
technology is also used to augment data from a small
number of tropical cyclone samples to generate the
training set.
 Train and validate the models using the collected data,
and evaluate their performance using appropriate metrics
such as accuracy, precision, recall, and error metrics
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 8
 Overall, the proposed research introduces novelty by
combining hybrid learning techniques, diverse data
integration, uncertainty quantification, and improved lead
time in the context of tropical cyclone intensity prediction.
 By addressing these aspects, the research aims to advance
the field and contribute to more accurate and reliable
predictions, ultimately benefiting disaster management and
coastal community resilience.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 9
 P Varalakshmi, N Vasumathi, R Venkatesan, Tropical Cyclone intensity prediction based on hybrid
learning techniques, Journal of Earth System, volume 132, Article number: 28 (2023) Springer.
 Xin Wang, Wenke Wang and Bing Yan, Tropical Cyclone Intensity Change Prediction Based on
Surrounding EnvironmentalConditions with Deep Learning,Water 2020, 12, 2685.
 Gao S, Zhang W, Liu J, Lin I.I, Chiu L.S, Cao K, Improvements in Typhoon Intensity Change
Classification by Incorporating an Ocean Coupling Potential Intensity Index into Decision Trees,
Weather Forecast. 2016, 31, 95–106.
 Li Y, Zhang H, ShenQ, Spectral–Spatial Classification of Hyper spectral Imagery with 3D
Convolutional Neural Network. RemoteSens. 2017, 967.
 Lee J, Im J, Cha D, Park H, Sim S,TropicalCyclone Intensity Estimation Using Multi-Dimensional
Convolutional Neural Networks from Geostationary Satellite Data. Remote Sens. 2019, 12, 108.
 Chen B F, Chen B, Lin H T, Elsberry R.L, Estimating Tropical Cyclone Intensity by Satellite Imagery
UtilizingConvolutional Neural Networks.Weather Forecast. 2019, 34, 447–465.
 Dvorak V 1975 Tropical cyclone intensity analysis and forecasting from satellite imagery; Mon. Wea.
Rev. 103 420–430.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 10
 “Rice Quality Prediction using Convolution Neural Network”, IEEE
Conference, International Conference on Distributed Computing and
Electrical Circuits and Electronics (ICDCECE-2023), 29th & 30th April, 2023.
 “IoT and Image Processing based Smart Door Locking System”, IEEE
Conference, 2022International Conference on Automation, Computing and
Renewable Systems, (ICACRS), 2022, 13-15 Dec. 2022, Pages 291-295.
 “Advanced Deep learning-based Controller for Robust EV Speed
Management”, IEEE Conference, 2022 Sixth International Conference on IoT
in Social, Mobile, Analytics and Cloud, (I-SMAC), 2022, 10-12 Nov. 2022, Pages
598-602.
ThankYou.
INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 11

Cyclone Predication using Machine Learing

  • 1.
     Coastal regionsin India are very frequently hit byTropical Cyclones, which result in tremendous loss.  Its intensity prediction has been a challenging task because of drastic climatic changes over the past few years in the world.  Intensity of Tropical Cyclone is highly influenced by ocean, atmospheric and meteorological parameters which makes the task difficult to define the mechanism of Tropical Cyclone intensity prediction. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 1
  • 2.
     To develophybrid learning models that can improve the accuracy of tropical cyclone intensity predictions. By combining multiple machine learning or statistical techniques(using both empirical and numerical modelling techniques)  To develop models that can provide reliable and timely predictions of cyclone intensification well in advance. By improving the lead time and warning systems, researchers aim to help communities and authorities make better-informed decisions and take appropriate actions to mitigate the impact of tropical cyclones.  To quantify and address uncertainties associated with predictions. The objective is to provide not only point forecasts but also probabilistic estimates, which can help decision-makers understand the range of potential outcomes and associated risks. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 2
  • 3.
     Here, ahybrid deep learning model is built using historical observations collected from various sources to perform a data-driven prediction ofTropical Cyclone’s intensity using regression model.  This hybrid model utilizes Convolutional Neural Network architectures for feature extraction and machine learning models for regression.  Application and Real-Time Implementation: The study may explore the practical application of the developed hybrid learning models in real-time forecasting and operational settings.  This could involve implementing the models within existing weather forecasting systems or assessing their feasibility for integration into decision support systems used by meteorological agencies or disaster management INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 3
  • 4.
     Varalakshmi.P etal. investigated Tropical cyclone intensity prediction based on hybrid learning techniques.  Xin Wang et al. proposed Tropical cyclone intensity change prediction based on surrounding environmental conditions with Deep learning.  Gao S et al. explained Improvements in typhoon intensity change classification by incorporating an ocean coupling potential intensity index into decision trees. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 4
  • 5.
     Li Yet al. explored Spectral spatial classification of hyper spectral imagery with 3D convolutional neural network.  Lee J ability to precisely extract Tropical cyclone intensity estimation using multi-dimensional convolutional neural networks from geostationary satellite data.  Dvorak V et al. proposed Tropical cyclone intensity analysis and forecasting from satellite imagery. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 5
  • 6.
     The existingliterature provides a strong foundation for the proposed study on tropical cyclone intensity prediction using hybrid learning techniques.  By building upon the previous research, this study aims to contribute to the field by developing novel hybrid models, incorporating diverse data sources, addressing uncertainty, and improving lead time and warning systems.  The findings from this work can have significant implications for disaster management and community resilience in the face of tropical cyclones. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 6
  • 7.
     Tropical cyclonespose significant threats to coastal communities and require accurate intensity prediction for effective disaster preparedness and response.  This research proposal aims to explore the application of hybrid learning techniques in improving tropical cyclone intensity prediction.  By combining multiple algorithms and integrating diverse data sources, we aim to enhance forecast accuracy and provide valuable insights into cyclone dynamics. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 7
  • 8.
     Design anddevelop hybrid learning models using a combination of machine learning algorithms, statistical methods, and physical models. Image processing technology is also used to augment data from a small number of tropical cyclone samples to generate the training set.  Train and validate the models using the collected data, and evaluate their performance using appropriate metrics such as accuracy, precision, recall, and error metrics INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 8
  • 9.
     Overall, theproposed research introduces novelty by combining hybrid learning techniques, diverse data integration, uncertainty quantification, and improved lead time in the context of tropical cyclone intensity prediction.  By addressing these aspects, the research aims to advance the field and contribute to more accurate and reliable predictions, ultimately benefiting disaster management and coastal community resilience. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 9
  • 10.
     P Varalakshmi,N Vasumathi, R Venkatesan, Tropical Cyclone intensity prediction based on hybrid learning techniques, Journal of Earth System, volume 132, Article number: 28 (2023) Springer.  Xin Wang, Wenke Wang and Bing Yan, Tropical Cyclone Intensity Change Prediction Based on Surrounding EnvironmentalConditions with Deep Learning,Water 2020, 12, 2685.  Gao S, Zhang W, Liu J, Lin I.I, Chiu L.S, Cao K, Improvements in Typhoon Intensity Change Classification by Incorporating an Ocean Coupling Potential Intensity Index into Decision Trees, Weather Forecast. 2016, 31, 95–106.  Li Y, Zhang H, ShenQ, Spectral–Spatial Classification of Hyper spectral Imagery with 3D Convolutional Neural Network. RemoteSens. 2017, 967.  Lee J, Im J, Cha D, Park H, Sim S,TropicalCyclone Intensity Estimation Using Multi-Dimensional Convolutional Neural Networks from Geostationary Satellite Data. Remote Sens. 2019, 12, 108.  Chen B F, Chen B, Lin H T, Elsberry R.L, Estimating Tropical Cyclone Intensity by Satellite Imagery UtilizingConvolutional Neural Networks.Weather Forecast. 2019, 34, 447–465.  Dvorak V 1975 Tropical cyclone intensity analysis and forecasting from satellite imagery; Mon. Wea. Rev. 103 420–430. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 10
  • 11.
     “Rice QualityPrediction using Convolution Neural Network”, IEEE Conference, International Conference on Distributed Computing and Electrical Circuits and Electronics (ICDCECE-2023), 29th & 30th April, 2023.  “IoT and Image Processing based Smart Door Locking System”, IEEE Conference, 2022International Conference on Automation, Computing and Renewable Systems, (ICACRS), 2022, 13-15 Dec. 2022, Pages 291-295.  “Advanced Deep learning-based Controller for Robust EV Speed Management”, IEEE Conference, 2022 Sixth International Conference on IoT in Social, Mobile, Analytics and Cloud, (I-SMAC), 2022, 10-12 Nov. 2022, Pages 598-602. ThankYou. INTERVIEW FOR Ph.D. ADMISSION FOR JULY 2023 SESSION 11