Paasha: A Tale of Chances
Predicting Election Results using Statistics and Machine Learning
Election prediction has always been of interest to social scientists and public in general. In this talk, we discuss the prediction of the number of seats for major parties in Maharashtra during the 2019 parliamentary elections. Past parliamentary and assembly elections data are used to develop a model utilizing some statistical and machine learning tools. The model selection and parameter tuning are made based on the retrospective model performance for the 2014 elections. The final predictions are almost entirely in agreement with the actual results of the elections.
05 March, 2022
Time: 15:30 - 16:30
Event Type: Webinar
Registration: Free
Deadline: 01 March, 2022
​
​
Role of Data Science and Machine Learning in Managing Plant Invasion Under Climate Change
Speaker
​
Amiya Ranjan Bhowmick currently Assistant Professor in the Department of Mathematics in Institute of Chemical Technology, Mumbai, India. He received his Ph.D. degree from the University of Calcutta, Kolkata in Applied Mathematics. Prior to joining to Indian Statistical Institute, Kolkata for pursuing Ph.D., he worked as a Statistician in the Cytel Statistical Software Services Pvt. Ltd. His primary research interest includes the application of growth models in biological research, species distribution modelling etc. He has publications in leading international journals in the broad area of Mathematical Biology. Apart from the research work he is actively involved in teaching Mathematics at undergraduate level and Statistics at post-graduate level. He is also actively involved in different organizational activities bridging the gap between biology and mathematics through workshops and conferences.
Biological invasions represent one of the major environmental challenges for the conservation of global biodiversity and the continuation of ecosystem services and have a serious impact on human and animal health and economic development. A comprehensive database, named, Indian Alien Flora Information (ILORA) contains data related to ecology, biogeography, introduction pathway, socio-economy, and distribution alien vascular plant species. The data set is expected to assist a wide range of stakeholders involved in policy formulation, and decision-making related to Invasive Alien Plant Species (IAPS). In recent years, several modelling tools have been used to predict the consequences of climate change on the spatial distribution of animal species, weeds, and crops in India. In this talk, I plan to discuss how the ILORA database will be useful in implementing policy regulations with the aid of data science and machine learning.
10 December, 2022
Time: 15:00 - 16:00
Event Type: Webinar
Registration: Free
Deadline: 07 December, 2022
​
​
What is genetic risk score and why is it important in precision medicine?
Speaker
​
Arunabha Majumdar is an Assistant Professor at the Department of Mathematics, Indian Institute of Technology Hyderabad (IITH). Prior to joining IITH, he worked as a postdoctoral researcher at the University of California, San Francisco and the University of California, Los Angeles. He is interested in various research areas including Statistical Genetics, Biostatistics, and Computational Statistics. He has developed several R-packages and published numerous papers in premier international journals including Nature Communications, Nature Genetics, and Biometrics.
Genetic studies have grown rapidly to better understand the complex genetic architecture of various life-threatening diseases like cancers. Statistics and computer science have played a pivotal role in the success of such studies. The genetic risk score (GRS) is defined to be an aggregate-level score combining the risk genetic variants and has the potential of predicting the future risk of an individual to develop a complex disease. Hence, GRS has started to play a crucial role in precision medicine, that is, a more targeted intervention strategy can be devised based on the genetic profile of a specific individual. This talk mainly focuses on the statistical foundations of GRS and how it can be applied to a specific disease.
18 December, 2021
Time: 15:30 - 16:30
Event Type: Webinar
Registration: Free
Deadline: 15 December, 2021
​
​
Good Graphical Principles and Innovative Graphs in Clinical Trials
Speaker
​
Pritam Gupta is currently working as a lead statistician for trials in rare disease and have also contributed to a vaccine study as the blinded safety statistician in Pfizer. Previously, he was an employee at Novartis, where he supported multiple therapeutic areas in the Global Medical Affairs group as a trial, lead and project statistician. He has contributed to planning and execution of exploratory analyses and statistical consultations, which has led to multiple publications in peer reviewed journals. He is involved in multiple teaching initiatives targeted towards non-statisticians and medical practitioners. He has received MS in Statistics from the University of Calcutta, India and a PhD in Statistics from the University of Wyoming, USA.
Quality visualizations are becoming very critical in today’s medical research for appropriate interpretation and decision making. Appropriate graphic design plays a critical role in the creation of graphs that efficiently and effectively translate the key clinical messages in the data. An effective graph is something that is easy for the audience to decode and one should bear in mind some key aspects which are influential and comprehensive for a quality visualization. The talk will focus on some of the best practices for guidance on visualization with clinical trial data along with discussion on some innovative plots created in the emerging area of visualization where effective graphics are becoming a fast and powerful tool to communicate key findings.
7 August, 2021
Time: 15:00 - 16:00
Event Type: Webinar
Registration: Free
Deadline: 4 August, 2021
​
​
Predicting Election Results using Statistics and Machine Learning
Speaker
​
Akanksha S Kashikar is an Assistant Professor at the Department of Statistics, Savitribai Phule Pune University. She received several awards including the Young Statistician Award from the Indian Society for Probability and Statistics, and the Young Scientist Award from the Society of Statistics and Computer Applications, India. She is interested in various research areas including Branching Processes, Count Data Time Series and Applied Statistics as well as interdisciplinary projects on biodiversity and astrophysics. She has published several papers in various international journals. She has a special interest in science popularization and has written articles related to Math-Stat in Marathi newspapers and periodicals. Recently, she composed a song on Regression Analysis, which also won an award from the Centre for Science Education and Communication, SPPU.
Election prediction has always been of interest to social scientists and public in general. In this talk, we discuss the prediction of the number of seats for major parties in Maharashtra during the 2019 parliamentary elections. Past parliamentary and assembly elections data are used to develop a model utilizing some statistical and machine learning tools. The model selection and parameter tuning are made based on the retrospective model performance for the 2014 elections. The final predictions are almost entirely in agreement with the actual results of the elections.
05 March, 2022
Time: 15:30 - 16:30
Event Type: Webinar
Registration: Free
Deadline: 01 March, 2022
​
​
Application of Statistical and Machine Learning Models in Complete Banking Customer Life Cycle
Speaker
​
Prakash Bade is currently an Associate Director at CRISIL Limited, London, United Kingdom. He has 12 years of experience in the data analytics industry with a master degree in Statistics from University of Pune, India. His major experience comes from Credit Risk analytics and Machine learning (ML) Modelling side. He worked with multiple global banking clients on Regulatory and non regulatory credit risk model development, validation and monitoring assignments. He got sound experience in setting up and leading small and large scale analytics teams. He is actively involved in academic collaboration, training, mentoring , and writing practitioner research papers in ML/AI and Credit Risk. He is an avid speaker at various public platforms and education institutes on the topics such as application of Statistics, Machine Learning (ML), Artificial Intelligence (AI) and credit risk analytics.
The Banking Industry has always been dependent on data driven decisions starting from customer acquisition, account management, compliance, risk mitigation and collection analytics. Given the current digital era and rise in technology adoption, real time analysis for risk reward optimization is taking new heights in the banking industry with the help of ML/AI models. Every stage of the banking customer life cycle is purely dependent on data driven decisions. This session will focus on evolving trends in Banking analytics, real life use cases of statistical and ML/AI modelling in the complete customer life cycle. This session will give a good understanding of application of ML/AI and Statistical modelling in Banking analytics and career opportunities in banking risk analytics.
10 September, 2022
Time: 15:00 - 16:00
Event Type: Webinar
Registration: Free
Deadline: 07 September, 2022
​
​
How to use different statistical models to analyze and predict COVID-19 in India?
_edited.jpg)
Speaker
​
Palash Ghosh is an alumnus of the Indian Statistical Institute and is currently attached with the Department of Mathematics, Indian Institute of Technology Guwahati as an Assistant Professor. He is also an adjunct Assistant Professor at the Centre for Quantitative Medicine, Duke-NUS Medical School, National University of Singapore. He is interested in various research areas including Dynamic Treatment Regime, Personalized Medicine, Sequential Multiple Assignment Randomized Trial, Design and Analysis of Clinical Trials, and Pharmacovigilance. His recent work on the analysis of COVID-19 data attracted attention from the Minister of Human Resource Development and various media houses.
COVID-19 is the single biggest challenge that the world has faced in recent times. For better management of COVID-19, we need to analyze and predict its course. This webinar will discuss different statistical models like Logistic, Exponential, and Susceptible-Infected-Susceptible (SIS) to predict COVID-19 in India. The emphasis is to interpret the results jointly from all models rather than individually. I will discuss the prediction issues that are specific to India compared to other parts of the world. The importance of Daily Infection Rate (DIR), a parameter-free measure, will be discussed in this context. In the end, I will introduce a dynamic data-driven algorithm to estimate the model parameters of a modified SIS model.
1 May, 2021
Time: 15:00 - 16:00
Event Type: Webinar
Registration: Free
Deadline: 15 April, 2021
​
​
Data Science in Action: Applications in Industry
Speaker
​
Suniti Srivastava is an alumnus of Indian Statistical Institute, with Masters in Statistics and specialization in Advanced Probability. She has worked for 14 years in the finance domain with leading companies like American Express, GENPACT and Royal Bank of Scotland. She has experience in various domains of finance like insurance, credit card risk management and investment banking. Her field of interest is Machine learning and its real world applications.
Statistics and probability theory is the science of decision making in the face of uncertainty. Statistics serves as a key pillar of support and guiding factor in decision making for most of the industries including retail, financial, tech, healthcare, telecom, and manufacturing. The foundations of machine learning and artificial intelligence are based on statistics and mathematics. This talk will focus on popular applications of machine learning techniques for default prediction and customer acquisition problems in the banking and telecom sector. Also, applications of stochastic models and analysis in risk management and Google page rank algorithm will be discussed.
16 January, 2021
Time: 16:00 - 17:00
Event Type: Webinar
Registration: Free
Deadline: 8 January, 2021
​
​