BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Asia/Kolkata BEGIN:STANDARD DTSTART:19451014T230000 TZOFFSETFROM:+0630 TZOFFSETTO:+0530 TZNAME:IST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20260317T091934Z UID:18FC6FA1-6E0A-4993-9C18-F194E24B18BD DTSTART;TZID=Asia/Kolkata:20260227T143000 DTEND;TZID=Asia/Kolkata:20260227T153000 DESCRIPTION:Ramaiah University recently hosted a distinguished guest lectur e by Dr. Asit Kumar Das\, Professor at the Department of Computer Science and Technology\, IIEST Shibpur. His talk\, “From SMOTE to VAEs: Sampling \n\nStrategies for Imbalanced Learning”\, explored the critical challeng e of class imbalance in machine learning\, particularly in domains such as disease detection\, fraud analysis\, and fault prediction. Dr. Das highli ghted the limitations of traditional models that often favor majority clas ses\, leading to poor minority class recognition. He discussed classical r esampling methods like SMOTE\, Tomek Links\, and hybrid approaches (SMOTE- ENN\, SMOTE-Tomek)\, before moving to advanced deep learning techniques su ch as GANs and Variational Autoencoders (VAEs). Special emphasis was place d on VAEs\, which provide stable training and probabilistic latent space m odeling\, making them effective for generating diverse synthetic\nsamples in complex datasets like medical imaging. The lecture also showcased exper imental results on COVID-19 chest X-ray classification\, demonstrating how latent space resampling with VAEs can improve recall and precision. Dr. D as concluded by stressing the importance of hybrid sampling and the potent ial of VAE-GAN combinations for future research. This insightful session e nriched students and faculty with cutting-edge strategies for tackling rea l-world imbalanced learning problems.\n\n[]\n\n[] []\n[] []\n\nSpeaker(s): Asit Kumar Das\n\nRoom A307\, Department of Computer Science and Engineer ing\, M. S. Ramaiah University of Applied Sciences\, University House\, Ne w BEL Rd\, M S R Nagar\, Mathikere\, Bengaluru\, Karnataka\, India\, 56005 4 LOCATION:Room A307\, Department of Computer Science and Engineering\, M. S. Ramaiah University of Applied Sciences\, University House\, New BEL Rd\, M S R Nagar\, Mathikere\, Bengaluru\, Karnataka\, India\, 560054 ORGANIZER:avamsikrishna@ieee.org SEQUENCE:10 SUMMARY:Guest Lecture: From SMOTE to VAEs – Sampling Strategies for Imbal anced Learning URL;VALUE=URI:https://events.vtools.ieee.org/m/545501 X-ALT-DESC:Description: <br /><p>Ramaiah University recently hosted a disti nguished guest lecture by Dr. Asit Kumar Das\, Professor at the&nbsp\;Depa rtment of Computer Science and Technology\, IIEST Shibpur. His talk\, &ldq uo\;From SMOTE to VAEs: Sampling</p>\n<p>Strategies for Imbalanced Learnin g&rdquo\;\, explored the critical challenge of class imbalance in machine learning\,&nbsp\;particularly in domains such as disease detection\, fraud analysis\, and fault prediction.&nbsp\;Dr. Das highlighted the limitation s of traditional models that often favor majority classes\, leading to poo r&nbsp\;minority class recognition. He discussed classical resampling meth ods like SMOTE\, Tomek Links\, and hybrid&nbsp\;approaches (SMOTE-ENN\, SM OTE-Tomek)\, before moving to advanced deep learning techniques such as&nb sp\;GANs and Variational Autoencoders (VAEs). Special emphasis was placed on VAEs\, which provide stable&nbsp\;training and probabilistic latent spa ce modeling\, making them effective for generating diverse synthetic<br>sa mples in complex datasets like medical imaging.&nbsp\;The lecture also sho wcased experimental results on COVID-19 chest X-ray classification\, demon strating how&nbsp\;latent space resampling with VAEs can improve recall an d precision. Dr. Das concluded by stressing the&nbsp\;importance of hybrid sampling and the potential of VAE-GAN combinations for future research.&n bsp\;This insightful session enriched students and faculty with cutting-ed ge strategies for tackling real-world&nbsp\;imbalanced learning problems.< /p>\n<p style="text-align: center\;" data-start="1340" data-end="1661"><im g src="https://events.vtools.ieee.org/vtools_ui/media/display/e1ec6eae-911 e-42f9-9fe7-6f514cff5d16" alt="" width="810" height="456"></p>\n<table sty le="border-collapse: collapse\; width: 100%\;" border="1"><colgroup><col s tyle="width: 49.9513%\;"><col style="width: 49.9513%\;"></colgroup>\n<tbod y>\n<tr>\n<td style="text-align: center\;"><img src="https://events.vtools .ieee.org/vtools_ui/media/display/1c35c71b-42be-4c91-903b-0bd88bd4fadc" al t="" width="400" height="239"></td>\n<td style="text-align: center\;"><img src="https://events.vtools.ieee.org/vtools_ui/media/display/bc45a7fb-5d98 -4b74-8662-bd563216e037" alt="" width="400" height="376"></td>\n</tr>\n<tr >\n<td style="text-align: center\;"><img src="https://events.vtools.ieee.o rg/vtools_ui/media/display/0efc696e-26d9-4e92-a654-8000e7e24c85" alt="" wi dth="400" height="371"></td>\n<td style="text-align: center\;"><img src="h ttps://events.vtools.ieee.org/vtools_ui/media/display/a22f5b57-3b3d-40e8-b 40a-cae0fdb274be" alt="" width="400" height="239"></td>\n</tr>\n</tbody>\n </table> END:VEVENT END:VCALENDAR