Prof. G. Raghuram

Topic: Data Analytics Opportunities for Indian Railways

Talk Abstract

This talk will focus on improved customer service and decision making through data analytics for Indian Railways. We start by characterizing where is the Indian Railways (IR) in terms of performance, both in terms of output and input parameters. We then identify the key challenges. In this context, we see what are the major data sets available with the IR. Based on this, we explore how can we use Data Analytics in improving the effectiveness of decision making related to output parameters in the context of passenger and freight businesses, and input parameters in the context of asset utilization and design.

Bio

Raghuram has been Director, IIM Bangalore, since February 2017. Prior to taking over as Director of IIM Bangalore, he was Professor and Chairperson of the Public Systems Group at IIMA. He has been Dean (Faculty), IIMA, Vice-Chancellor of the Indian Maritime University and Indian Railways Chair Professor. He specializes in infrastructure and transport systems, and logistics and supply chain management. He conducts research on the railway, port, shipping, aviation and road sectors. He has published over 35 refereed papers in journals and written over 160 case studies. He has published six co-authored books. He was awarded (i) ‘Life-time Achievement Award’ for Transport Excellence, by Mahindra and Mahindra, supported by Ministry of Road Transport and Highways in 2018 (ii) ‘MC Puri Memorial Award’ for contribution to Operational Research in India in 2016 (iii) ‘Lifetime Achievement Award’ for contribution to Logistics and Infrastructure by EXIM News in 2014 and (iv) ‘Academician of the Year’ by the Chartered Institute of Logistics and Transport in 2012. He is a Fellow of the Operational Research Society of India, and Chartered Institute of Logistics and Transport. He has teaching experience at universities in India, USA, Canada, Yugoslavia, Singapore, Tanzania, UAE and Japan.
He is a member of the Global Future Council on Mobility of the World Economic Forum and of the Board of Directors of three companies in the fields of infrastructure and logistics. Overall, he has been on the Board of 13 companies. He is on the Board of four educational institutions, one NGO and on the Advisory Board of a few other educational institutions. He has offered consultancy services to over 100 organizations including multilateral agencies. He has been part of various government policy making and advisory committees for the Ministry of Civil Aviation, Ministry of Consumer Affairs and Public Distribution, Ministry of Railways, Ministry of Road Transport and Highways, Ministry of Shipping, Cabinet Secretariat, Comptroller and Auditor General, the Planning Commission and various State Governments.
Raghuram has a BTech from IIT, Madras; a Post Graduate Diploma in Management from IIM, Ahmedabad; and a PhD from Northwestern University, USA.

Prof. R.K. Amit

Topic: Fermat meets Shapley—Location Science and Game Theory

Bio

Dr. R K Amit is currently an Associate Professor in the Department of Management Studies, IIT Madras. He completed his undergraduate studies at IIT Kanpur, and his doctoral studies at IISc, Bangalore. His research and teaching interests are game theory and decision theory, and their applications in operations management. His research has been published in top-tier journals like European Journal of Operational Research, Operations Research Letters, International Journal of Production Economics, Omega, and Journal of Cleaner Production. He is currently working on numerous industry- sponsored research projects in the areas of electric mobility, emergency medical services, and airlines revenue management.

Talk Abstract

Location science focuses on finding site for one or more activities to optimize one or more objectives. Location science covers the problems of land allocation, the location of industrial and communication facilities, and the location of healthcare and emergency services.
The quantitative origins of location science dates back to the early 1600s, when Pierre de Fermat discussed the median location problem. One of the first game theoretic models for location problems was developed by Harold Hotelling in 1929 to compute an equilib- rium location for two competiting vendors. Since then, game theoretic models have been enriching location science. In this talk, I will focus on the interface of location science and game theory, including our research on using the Shapley value from cooperative games for ambulance repositioning.

Prof. Deepak Subramani

Topic: Optimal Path Planning in Dynamic Environments

Bio

Deepak Subramani is an Assistant Professor in the Dept. of Computational and Data Sciences at IISc, Bangalore. His research is at the intersection of computation, stochastics and geosciences, with a focus is to build holistic science-based data-driven computational solutions to complex engineering and environmental problems. Previously, he completed his postdoc, Ph.D. and MS from Massachusetts Institute of Technology, and Dual Degree (B.Tech+M.Tech) from IIT Madras. His recent research projects and publications can be found on his webpage: here

Talk Abstract

Optimal planning of autonomous marine platforms is extremely important for their efficient use in a variety of scientific, security and humanitarian applications. In this context, we develop fundamental stochastic partial differential equations and efficient solution schemes for time, energy, risk-optimal paths of agents in strong, dynamic and uncertain currents. We apply our schemes and software to plan paths in a variety of idealized, realistic and real time operations of AUVs, gliders and ships in different coastal ocean regions. Notably, our planning rigorously incorporates environmental predictions and fundamental optimal control PDEs in a probabilistic framework.

Prof. S.K. Neogy

Topic: Recent Advances on Complementarity Problems and its applications in Stochastic games

Bio

S. K. Neogy is currently a Professor at Indian Statistical Institute Delhi Centre. The main focus of his research in the last two decade has been to explore problems on Complementarity Problem and special matrices that arises in Nonlinear programming, Non-cooperative games, Stochastic Games and various other applications. He has coauthored, coedited a number of research monographs and special issues of journals. More details are available at: https://www.isid.ac.in/~skn/. The focus of his current research is characterization of matrix classes using degree theory, introduction of new matrix classes that extend processability of Lemke’s algorithm and Cottle-Dantzig algorithm, finite step algorithms for structured Stochastic games and Semi-Markov games.

Talk Abstract

Complementarity problem is well studied in the literature on Mathematical Programming and arises in many applications in Operations Research, Mathematical Economics, Engineering and Game theory. The linear complementarity problem has played an important unifying role in operations research since its introduction for more than five decades ago. The distinguishing feature of a complementarity problem is the set of complementarity conditions. A major source of complementarity problems arises from optimality conditions of general constrained optimization problems. The linear complementarity problems are found in the first-order optimality conditions of quadratic programming. Probably the earliest publication containing an explicitly stated linear complementarity problem is one by Du Val (1940) in algebraic geometry. Another important area of applications of complementarity problems is in financial mathematics. For example, the linear complementarity problem has been successfully applied to option pricing. Several optimization problems like linear programming, linear fractional programming, convex quadratic programming, the bimatrix game problem, market equilibrium, optimal capital stock, optimal stopping problems and some classes of structured stochastic game problem are presented in a unified framework called complementarity problems. It is always an extremely demanding and interesting topic to researchers in the field of optimization.

Prof. Raghu Nandan Sengupta

Topic: Reliability and Robust Portfolio Optimization: An Introduction

Talk Abstract

Financial markets are being increasingly exposed to both exogenous as well as endogenous uncertainties. Hence formulating optimal portfolios, with accurate estimates of risk exposure, is a major challenge in asset allocation problem. We discuss an innovative solution technique called reliability based robust optimization (RBRO) method for asset allocation problems, where both asset returns (hence parameters of asset returns) and input data are probabilistic. Furthermore copula concept along with extreme value theory (EVT) is incorporated to tackle these problems.

Prateek Jain

Topic: Resource Efficient ML in 2KB of RAM

Talk Abstract

Several critical applications require ML inference on resource-constrained devices, especially in the domain of Internet of Things like smartcity, smarthouse etc. Furthermore, many of these problems reduce to time-series classification. Unfortunately, existing techniques for time-series classification like recurrent neural networks are very difficult to deploy on the tiny devices due to computation and memory bottleneck. In this talk, we will discuss two new methods FastGRNN and EMI-RNN that can enable time-series inference on devices as small as Arduino Uno that have 2KB of RAM. Our methods can provide as much as 70x speed-up and compression over state-of-the-art methods like LSTM, GRU, while also providing strong theoretical guarantees.

Pankaj Dayama

Topic: Fair Division using Blockchain

Bio

Pankaj Dayama is a Senior Researcher and Manager at IBM Research. He is currently leading Blockchain Solutions group at IBM Research with Privacy Preserving Network Collaboration as one of the key focus area. His group also works directly with IBM clients to build innovative blockchain solutions.

Talk Abstract

Fair division techniques provide axiomatic solutions for sharing finite resources (or cost or value) in horizontal supply chain collaborations. We explore in this talk the use of permissioned blockchains to implement arbiter-less fair division mechanisms. The smart contracts and distributed ledger capabilities of blockchain provide the necessary computational and consensus environment to implement fair division techniques without a mediator. We showcase with examples (collaborative shipping, inventory pooling, rent division, etc) the optimization and algorithmic challenges in blockchain implementations due to privacy issues.

Prof. Vikas Vikram Singh

Topic: Stochastic Nash games under chance constraints

Bio

Dr. Vikas Vikram Singh received his PhD degree in Industrial Engineering and Operations Research from IIT Bombay in 2013. After his PhD he spent one year at INRIA, Sophia Antipolis, France, and two years at Universite Paris Sud as a postdoctoral researcher. He is mainly interested in game theoretic problems involving distributional uncertainties. In particular, he has worked on research problems from areas of stochastic dynamic games, network formation games, and chance constraint based stochastic Nash games. His current focus is on stochastic Nash game problems using the tools from stochastic optimization and distributionally robust optimization. He is currently an Assistant Professor in the department of Mathematics at IIT Delhi.

Talk Abstract

The strategic games with stochastic payoffs and constraints have been extensively studied in the literature. The case of risk neutral decision makers has been studied by considering the expectation of random payoffs and constraints. We use chance constraint programming to model the risk averse stochastic Nash games. We show the existence of Nash equilibrium for the case of elliptically symmetric distributions. We also consider the case where underlying probability distribution is only known to belong to a distributional uncertainty set. These cases are modelled by considering the worst case situation using distributionally robust approach. We showed the existence of Nash equilibrium for various distributional uncertainty sets. We also discuss some computational approaches to compute the Nash equilibrium for these games.

Prof. P Balamurugan

Topic: Classifying bio-medical images using induced deep Region of Interest extraction - Case studies on Retinal and Cardiac images

Talk Abstract

Deep neural networks are known for their impressive performance in object detection and classification in "images in the wild" (images obtained using a simple internet search). However, only limited success has been observed in bio-medical images, especially those obtained from Magnetic Resonance Imaging (MRI) scans. In this talk, I will present a framework for classification and regression problems in bio-medical images. Our approach extends the popular U-deep nets to facilitate learning using an induced region of interest extraction. I will present some results obtained on retinal and cardiac images. This is an ongoing work with Ashutosh Kushwaha and Jaswant Singh.

Dr. K.Ganesh

Topic: Solving MEIO as Spanning Tree Problem using DP

Bio

Dr.K.Ganesh is Senior Knowledge Expert and Global Lead of “manufacturing and supply chain management center of competence” at McKinsey & Company, Chennai, India. He has graduated in Production Engineering from Annamalai University with the university first rank. He pursued his Masters’ degree in Industrial Engineering at NIT Trichy and secured college first. He then moved to IIT Madras where he obtained his doctoral degree in Supply Chain. He is having total 14 years of research and consulting experience in the supply chain consulting and worked with top consulting firms such as TCS and IBM. He has published more than 120 papers in leading international journals and written 12 international books. He is editors for 2 international journals and in reviewer board for 10+ renowned international journals. He is himself a lyricist and has written several poems in both English and Tamil.

Hemant Kowshik

Topic: Optimization and Predictive modeling in Computational Advertising

Talk Abstract

Advertising at Amazon is a fast-growing multi-billion dollar business that spans across desktop, mobile and connected devices; encompasses ads on Amazon and a vast network of hundreds of thousands of third party publshers; and extends across US, EU and an increasing number of international geographies. Computational Advertising is one of the most challenging areas for algorithmic optimization due to the scale of the problem, direct impact on the business and because of the interplay of multiple areas like machine learning, statistics, data mining, data streams, computational economics and econometrics. In this talk, we will provide an overview of a few problems in the advertising spectrum ranging from robotic traffic detection to publisher quality scoring to user engagement prediction to bid optimization and pricing. We describe the challenges and business value in building advanced and highly scalable algorithms to optimize performance for advertiser, publisher, ad-network and user.

K Ravikumar

Topic: Data-Driven Approaches to Improving Operational Efficiency in Drilling

Bio

Dr. Ravikumar is currently the Director of AI & ML at SigmaStream, a Texas-based start-up that develops digital technologies for the Oil and Gas industry.

Prior to joining SigmaStream, he was a Principal Scientist in Digital Manufacturing and Operations at the TCS Innovation Labs, Cincinnati/Bangalore. He held the Chief Statistician position at Manthan Systems, Bangalore from 2012 to 2014 and led the technical effort on Offer Recommendation System module of Manthan's customer analytics product, TargetOne. He was a researcher at the General Motors India Science Lab- Bangalore from 2003 to 2012, heading the Enterprise Analytics group. He worked as a Research Staff Member at the IBM India Research Lab, New Delhi during 1998 - 2003.

He received the Charles McCuen Special Achievement Award, and Inventor MIlestone Award from General Motors Global R&D. His work at TCS on Enterprise System Response Time Prediction won the best PoC award at the Auto Digital Day- Mumbai 2017.

Ravikumar's research interests are in the areas of supply chain management, revenue management, retail analytics, drilling optimization, and medical image analysis.

Ravikumar received his Ph.d from the Department of Computer Science and Automation, IISc, Bangalore, Masters in Industrial Engineering from NIT Calicut, and B Tech in Mechanical Engineering from Andhra University, Visakhapatnam.

Talk Abstract

Several sensor measurements are collected from drilling rig during oil well drilling process. While surface measurements can be collected in real-time, downhole measurements can only be obtained after a time lag. These measurements carry information about the operational states of the drilling. Real-time information about operational states will help the drilling crew drive operational efficiency. Automatic detection and classification of drilling operational states is considered as a big challenge in the drilling industry. Starting with a brief introduction to oil well drilling, this talk will primarily focus on our data-driven approaches to operational state detection and will shed some light on a few important research challenges in controlling a drilling system for optimal operational performance.

Prof. Ankur Sinha

Topic: Bilevel Optimization: Methods and Applications

Talk Abstract

Bilevel problems are special kind of optimization problems that involve two levels of optimization, namely upper level and lower level. The hierarchical structure of the problem requires that every feasible solution to the upper level problem should satisfy the optimality conditions of the lower level problem. Such a requirement makes bilevel optimization problems difficult to solve. These problems are commonly found in many practical problem solving tasks, which include optimal control, process optimization, game-playing strategy development, transportation problems, coordination of multi-divisional firms, machine learning and others. Due to the computational expense and other difficulties involved in handling such problems, they are often handled using approximate solution procedures. There is a need for theoretical as well as methodological advancements to handle such problems efficiently. In this talk, I will introduce bilevel optimization applications and methods that are used to handle such problems. Some recent ideas that researchers are working on towards solving such problems efficiently will also be highlighted..

Sudarshan Ravi, LetsTransport

Topic: Moving from HI to AI

Bio

Sudarshan is a 2013 batch dual degree from the Indian Institute of Technology, Kharagpur. He founded LetsTransport in 2015, a last mile logistics company, where he is responsible for product and strategy. Over the last 4 years, LetsTransport has grown to over 9 cities catering to a driver network of over 25,000 trucks. He has been instrumental in building the team at LetsTransport and creating the new-age product portfolio of the company to cater to some of the biggest enterprises in India. Prior to LetsTransport, Sudarshan spent his career working as an investment professional at JP Morgan Investment Bank.

Talk Abstract

How embedded is AI in our daily lives? Is AI growing meaningful in the digital age with the exponential growth in data being produced every day? Can AI replace HI? Will the man-machine combination boost productivity? How are new age companies leveraging AI to solve problems for masses? In this talk, the speaker will address the above questions by bringing into light how AI has been helpful in enabling growth and scale at LetsTransport. He will also briefly touch upon the various projects around clustering, fraud detection, forecasting, and operations control running at LetsTransport.

Dr. Deepak Paramashivan

Bio

Dr. Deepak Paramashivan had his training in Sarangi from Mr. Faiyaz Khan, and underwent advanced training under maestros Pandit Ramnarayan, Grammy winner, late Pandit Dhruba Ghosh and currently he is under the tutelage of world renowned Tabla and Sitar maestro Pandit Nayan Ghosh. He has performed in Europe, USA, Canada, UK, New Zealand, Singapore, South Korea both as a soloist and collaborating with world renowned artists such as Ustad Aashish Khan, Pandit Anindo Chatterjee, Pandit Yogesh Samsi, Abbos Kosimov and Pandit Birju Maharaj. He has played the Sarangi for A R Rehman's Roli seaboard and for the Hollywood music director Rick Boston's upcoming movie Messiaen. He is a recipient of the Killam Award, Andrew Stewart Award and the Edmonton Arts Council award from Canada. Deepak has a PhD in energy and climate engineering, with a gold medal, from the Indian Institute of Science. After serving a stint at the University of Alberta as an NSERC climate scientist, he switched gears completely and he recently completed his second PhD Ethnomusicology at the university of Alberta, Canada.

Ishaan Ghosh

Bio

All of 18, Ishaan Ghosh stands out as one of the front-ranking Tabla players in the younger generation of India’s Classical Music scene. He is the grandson of one of the most acclaimed Tabla Maestros of the 20th century, Late ‘Padmabhushan’ Nikhil Ghosh and the son and disciple of the internationally renowned Tabla and Sitar stalwart Nayan Ghosh.

An exceptionally gifted child prodigy, Ishaan gave his first Tabla Solo recital at the age of two and a half.

Since the age of 8, Ishaan has presented traditional Indian music at prestigious venues in Switzerland, France, Spain, USA, UK, Mexico, Hawaii among other countries.

His dynamic Tabla Solos have received wide appreciation from musicians and connoisseurs alike. Besides accompanying his father’s sitar concerts, he has also performed with landmark figures in the field of Indian Classical Music like ‘Sangeet Martand’ Pt. Jasraj, Ustad Amjad Ali Khan, Padmabhushan Pt Buddhadev Dasgupta, Ustad Aashish Khan, Pt. Tejendra Narayan Majumdar, violin legend Dr. N Rajam, Rakesh Chaurasia & Purbayan Chatterjee, among several other prominent artistes.

He has regularly collaborated with musicians of various other genres like Jazz Guitarist Max Clouth of Germany, Mohammad Rasouli and percussionists Fakhroddin Ghaffari and Hamid Reza Rahbaralam of Iran, Hoo Eouling of China and Raho Langsepp of Estonia.

He was conferred the ‘BABA ALLAUDDIN KHAN YUVA PURASKAR’ by the Government of Madhya Pradesh, India in 2012.

In 2016, he was specially invited along with his father by the former President of the United States, President Jimmy Carter to give a benefit concert for the HABITAT FOR HUMANITY project, and was also felicitated by the President himself.