Online conference in Data Science: Talks
Some days back I have attended a virtual conference in Data Sceince organized by ACM SIGKDD and IIT Madras.
This year I have started to attend few of conferences to get the latest researh trends in Data Sceience/AI research in India and globally. As I have recently started my Ph.D. from IIIT, Allahabad (check my earlier post “Why I have started PhD after working 15+ in Industry?”).
**AI & ML in the Times of COVID Prof. Partha P. Chakrabarti aka PPC, IIT Kharagpur
**A data-driven approach for country-level COVID Management Dr. Kamakoti V.
Arogya Setu case study – data driven model PostgreSQL, Cloud, Blutooth, epidemolists
Mis-information/fake news Joyojeet Pal. Microsoft
**City-scale agent-based simulator for modelling COVID-19 spread
Video Prof. Rajesh Sundaresan, IISC, Banglore Agent based Simulator and model for Covid 19 how the covid spraed in close intration, how agent based works in citi based system , study done for blore and mumbai. and good case study of proposed school opening in balore. trains restart, school restarts traffic modelling, odd-even in delhi
**The State and Fate of Linguistic Diversity and Inclusion in the NLP World Dr. Kalika Bali, Microsoft The State and fate of linguistic diversoty inclusion in NLP world, No of resources, no of people speak that language Video
**Object Detection in Scientific Plots Video Dr. Mitesh Khapra, IIT, Madras natural images vs sifitific images object detection, why object detetction is not sucessful localization accuracy
**Landscape of regularized Auto-Encoders for generative modelling: Intro, challenges & new directions
Video Dr. Prathosh A.P. This talk is bit mathematical and focusses more on new reasrch on adversial encoders etc how the Gerative model works,
**Sketch-based Image Retrieval
Video Soma Biswas Sketch retrival cross modal retirval SBIR ( sketch based image retrival) basic idea is to give sketch and get the corresponding image make a sketch and get a image in serach from a catalogue so input is different modality and output is diffent that’s called cross modal modality in surveliience match a sketch with criminal list with the database Zero shot leaaning not seen a single image in training