Advanced Analytics & Machine Learning on Cloud - Creating Impact for Education, Healthcare & more Date:   Tuesday , December 08, 2015 With the emergence of powerful distributed computing environment and on-demand cloud infrastructure, computing today is increasingly becoming affordable, scalable, and accessible to almost anyone, anytime and anywhere on this planet. The cloud platform now and in the future should be able to scale up or down to match demanding workloads. This democratization of computing has deep implications to how we set about solving complex social issues that involve processing and analysis of vast amounts of data to derive actionable insights in a timely manner. Advances in computing to such an extent can help governments and society create relevant interventions in pressing issues. Big data analytics, coupled with machine-learning algorithms, hold immense potential in terms of driving fundamental changes to critical areas like education and health. What governments need today to make a deeper social impact is an intelligent cloud that enables efficient processing of vast volumes of data with rich visualization & analytics capabilities. Deriving actionable insights on this data will help them make informed interventions in areas of concern. 10,000 schools in Andhra Pradesh to help reduce dropout rates with machine-learning Reducing the number of school dropouts and increasing the overall skill levels of the workforce are key areas of concern for local and state governments in India. An intelligent cloud can help drive skill levels to an appreciable level. It can do this by processing complex data on education, which includes data on student performance, school infrastructure, teacher skills, and enable experts to cull meaningful patterns from such data. This can lead to faster decision-making, replacing what would otherwise be a long drawn-out manual exercise. Further, it significantly reduces the chances of human error. The approach involves deploying machine-learning models that can predict the chances of a student dropping out. Microsoft has engaged with the Government of Andhra Pradesh from several months to help harness Azure-based machine-learning and advanced analytics tools to provide meaningful insights to the government aimed towards reducing the rate of school dropouts. This will help the AP government to design and implement targeted intervention strategies. This is a first-of-its-kind initiative by any state government in India to perform this sort of complex parsing of data on public schools and deploy machine-learning technology to predict dropouts. This cloud-based solution, using machine learning models, is now able to predict student dropouts quite accurately and is currently live on Azure for over 10,000+ government schools across the entire state of AP. So far, more than six lakh predictions have been generated from this initiative and this has enabled the officials in the education department to make timely interventions in the \'Badi Pilosthundi\' program to prevent dropouts in a significant way. Advanced analytics gives new hopes to people with visual disabilities Globally, there are 285 million people who are visually impaired, with 54 million in India alone. In March 2015, Microsoft engaged with one of India\'s leading eye care institutes - LV Prasad Eye Institute (LVPEI) to see how advanced analytics and machine-learning could help the organization derive insights from clinical data and tackle eye diseases more effectively. LVPEI has treated more than 18 million patients since 1987, and in 2010, it deployed an on premise system to collect patient data - electronic medical records (EMR) across its 120+ centers in India. Azure has helped collect insights from millions of clinical data points from those 120+ centers across multiple states for patients suffering from eye diseases, along with a mapping of geographies where these occur. In a successful pilot program, a team of LVPEI doctors worked closely with Microsoft to build predictive models using cloud-based machine-learning. One such experimentation recently done was in the area of Lasik surgery - to predict the regression rate. The idea was to use past clinical data and the correlations to predict the risk factor for the patient and make it available to the doctor prior to surgery. The model was built to predict what would be the new \"eye number\" (Uncorrected Visual Acuity or UCVA) 1 day/1 week after the surgery. This is aimed at predicting surgery outcomes, such as probability of success (or failure), and could well be a cost-saving measure for patients. Machine-learning & advanced analytics in various areas of human resource development and sectors like technology can indeed play a decisive role in addressing many of our current concerns and technology players must work closely with governments to create far-reaching social impact. Above all, the coming age of affordable advanced computing and its democratizing effects, gives us new hopes for a better tomorrow.