Search Talks

Dive into the fascinating world of probability, statistics and data mining as we embark on a journey to unravel complex real-world challenges. Through captivating examples, stories and anecdotes, discover how these statistical principles enable us to quantify the unknown, navigate uncertainty, and devise effective strategies for addressing a wide range of real-world problems. We also discuss about the journey of statistics through time.

Generative AI has become a popular sensation, showing remarkable abilities including with language, analogical reasoning and originality. But, as is well known, it is notoriously unreliable.

Less visible have been dramatic developments in Formal Proof Systems, where computers verify and help in finding proofs.

Formal proof systems perfectly complement generative AI for Mathematical reasoning, at least in principle. In this talk, I discuss these two developments and their relation.

Indian women work ten times more than their counterparts in other countries. Yet, as per official statistics, only 1 out every 4 working age women in India are employed. There have been several arguments made to answer this question. One strand of the literature points towards the conceptualisation of the notion of employment and its limitations in identifying and recognising the kinds of work that women do. Relatedly, some argue that ways of data collection inherently fails to capture women's employment - 'data is sexist!', as they say. A second strand argues that women's domestic work, care burden and gender-based norms keep women out of the paid workforce. Finally, a more recent strand points towards the lack of job creation as a fundamental reason for fewer women in the workforce. Based on primary surveys and official secondary data, the talk addresses each of these strands to unpack the low and falling levels of women's employment in India. I draw from my personal experiences as a...

Fourier decomposition into component frequencies helps to describe audio signals in terms of musical attributes such as melody and timbre. We link the components to the perception of the sound and show the utility of the analysis in applications such as music search and source separation.

Seen to be at the vanguard of the STEM to STEAM movement, Science Gallery is the world’s first university-linked network dedicated to public engagement with science and art with galleries in Atlanta, Dublin, London, Melbourne, and Monterrary. Science Gallery Bengaluru is the first such gallery of the network in Asia with the mandate to deliver a global mission. In this talk, Jahnavi Phalkey will introduce the form and structure of the new institution, and explore with us the ideas that inform the shape of Science Gallery Bengaluru through her journey to answer one question: what does it mean to go beyond science communication towards an engagement with science, and what does history of science have to do with it?