Extreme weather events have become the norm of our day-to-day lives! We hear long running heat waves, flash floods, extreme droughts, fires, and failed monsoons. These extreme weather events have an unprecedented impact on psychological and quality of life on the population, especially the poor and underprivileged will bear most of the impact in terms of loss of economic opportunities and un-sustained livelihood. The result is increased undernourishment and food insecurity. The preparedness of future democratic governments and public distributed systems rest in harnessing prognostic markers from data using advanced analytics from past weather events. The well-prepared governments apply multi-modal interdisciplinary governmental data systems to perpetually analyze and infuse departments to be ready to tackle any potential disruptions to the lives of citizens, especially poor and marginalized parts of the societies that could fall into vicious cycle of poverty-health-and-loss of economic opportunities. The paper proposes innovative Machine Learning Models to address food security concerns to the extreme weather events.

The budget speech is part of the democratic process that is presented annually to members of the parliament and addressed to Speaker of the house. Budget speech includes details of annual financial statements or financial plans of the government, containing details of revenue and expenditure in the past, along with the estimated spending and projections for the following year. Speech, additionally, consists of new policies and / or reforms announced to address fiscal macroeconomic issues. It takes, importantly, years to witness effectiveness of policies, especially in agriculture and infrastructure sectors. In this research paper, we propose an innovative Machine Learning framework that scores effectiveness of agricultural policies through binning language processing statements with key macroeconomic performance multiclass-multilabel-indicators that are regressed from government finance statistics and macroeconomic time series data. Finally, the paper presents budget speech prototype solution as well as its application for analyzing 2021 Indian budget speech.

Microsoft Azure DevOps is a robust ,cross platform and powerful automation engine for script-based automation tools. Azure DevOPS enables to build, test, and deploy Cloud Native and/or Non-Cloud Native applications. The core principle and chief advantage that Azure DevOps provide are the availability of automation techniques such as infrastructure as code and the seamless integration of verifiable frameworks such as Machine Learning Operations (MLOps) with the DevOps automated pipelines to provision and configure the infrastructure that applications need to run.With the increase in application complexity and with the infusion of Machine Learning (ML) and Artificial Intelligence (AI) techniques as part of the software development lifecycle, the Azure DevOps is the most important framework that many organizations are rapidly progressing to incorporate it in their business processes to reduce the cost of building product and improve customer success.As part of the paper, we would like to propose a novel DevOps framework for building intelligent Tiny ML dairy agriculture sensors and the advantages that DevOps provide to develop high quality product in the most cost-efficient manner and serve small scale farmers who are at the bottom economic pyramid.

House calls have nostalgic view and have practiced decades ago when the doctor arrived at the patient's door carrying a big black bag. House calls, in Electronic Health Records (EHR) era, are performed by clinicians sifting through EHR diagnostic or encounter records that exhibit a health condition, an anomaly or a violation of health rule set by the primary physician. House calls could prove to be a better way of treating very sick, elderly patients while they can still live at home. One of the greatest benefits of house calls is avoidance of Healthcare associated infections, especially during the Coronavirus (COVID-19) epidemic. Prioritizing patients on to house call list in the shortest amount of time is one of the daunting challenges that many healthcare institutions are facing. The reasons could be growing data volume of healthcare patient cases with combinatorial possibilities of disease conditions intermingling with the COVID-19 pandemics paralyzing a human agent to prepare house call list on a daily basis. The solution is to employ optimization techniques powered by mathematical formulations and derive solution by running solvers to generate priority list of patients so that the healthcare providers have a greater coverage of their needed patients' house calls are performed in-time. In this paper, we propose innovative novel idea "mathematical formulation enabled house calls". Finally, as part of the paper, we will present Sanjeevani house call service that is been deployed and currently in production.

Climate change is challenging our way of lives. Rising global temperatures are triggering increases in the frequency and severity of extreme climatic events, such as floods, droughts, & heat waves and resulting in unprecedented increase in economic cost and human impact. For example, Global warming cost the U.S. government more than $350 billion between 2007 and 2017 and will cost $112 billion per year in the future, according to the U.S. Government Accountability Office. Not only economic impact, global warming, importantly, is increasing the human fatality rates. The environmental and health research clearly suggest the linkage between global warming and increase in heat-related mortality, particularly in low-latitude countries, such as India, where heat waves will become more frequent and populations are especially vulnerable to these extreme temperatures. A clear example is a heat wave, scorching temperatures reached a record 116 degrees Fahrenheit, that struck the Indian city of Ahmedabad in 2010 killing hundreds of its most vulnerable citizens. The World Health Organization (WHO) estimates the climate change contributes to 150,000 deaths each year. By 2030, that number will double. The dire consequence of global warming and increase temperature mortality rates is unequivocally faced by the most vulnerable population, Senior Citizens, due to lack of proactive insights and timely availability of healthcare services. We strongly urge to include global warming related extreme events and its impacts on patients in Electronic Health Records (EHR) and to be imperatively considered as patient well-being data. Next, Electronic Health Records continuously synthesize relationship between Electronic Health Records (EHR) patient health episodes/encounters due to extreme climate events, albeit globally in a de-identified manner, and application of active machine learning techniques: A priori and naive Bayesian and artificial intelligence (AI) to derive hierarchical cohort of high risk senior citizens’, outpatients, clusters and proactively delivering population health insights and delivery of timely health services through alerting governmental and non-governmental (NGO) agencies. The golden standard for success is to counter high temperature related mortality rates and to prevent deaths due to temperature related to global warming. In this research paper, we propose the development of high-risk temperature to high mortality cohort cluster-based Machine Learning (ML) / Artificial Intelligence (AI) algorithms to prevent deaths of senior citizens. Finally, the paper presents a cohort cluster prototype solution as well as its application and certain experimental results.

The report summarizes the various techniques and feature engineering processes that we have applied for the Far Eastern Memorial Hospital (FEMH) Voice Data Challenge. We have used Mel scaled spectrograms and MFCC components as audio features to train various Neural Network Architectures. We have trained a 5-layer plain network, 5-layer CNN and RNN. We discuss the challenges faced and solutions to improve model performance, model parameter tuning and model evaluation.

Selfies are popular. They embrace and represent social and emotional pulse of the User. We offer, nevertheless, groundbreaking and novel radical view on Selfies, especially Selfies that are taken for medical image purposes. In our view Selfies that are taken for medical image purposes are valuable outpatient healthcare data assets that could provide new clinical insights. Additionally, they could be used as diagnostics markers that could provide prognosis of a potential masked disease and necessitate actions to avert any emergency incidence, thereby saving Billions of dollars. We strongly believe that Interweaving Selfies that are taken for medical image purposes with outpatient Electronic Health Records (EHR) could breed new data driven diagnosis and clinical pathways that could potentially preempt healthcare services rendering decision making process for greater efficiencies and that could potentially save valuable time and attention of healthcare professionals who're already operating on a highly constrained time and shortage of skilled human resources. Putting in simple terms, Selfies could offer new diagnosis & clinical insights that have the potential to improve overall health outcomes of people around the globe in a cost-effective manner that epitomizes the confluence of popularity with curiosity and sharing with accountability.In this research paper, we propose computer vision (CV) based Machine Learning (ML) / Artificial Intelligence(AI) algorithms to classify and stratify Selfies that are captured for medical imaging purposes. Finally, the paper presents a CV - ML/AI prototyping solution as well as its application and certain experimental results.