Sandeep Pawar
Simulation & Data Analytics Engineer
Chicago, Illinois, United States
Sandeep is a Mechanical Engineer turned Data Analytics engineer, specializing in using Data Science to solve Manufacturing and Product Development challenges.
Area of Expertise
Topics
Time Series Forecasting in Power BI
Time Series Forecasting is essential for any size business to prepare and plan for the future. While Power BI has an in-built forecasting feature that is easy to use, it is severely limited, poorly documented and often misused. In this session, I will explain how Power BI forecasting works, its strengths/limitations, how to overcome those limitations and create an advanced forecasting model.
Power Up Power BI with Jupyter Notebook - Use Cases
Jupyter Notebook is used by data scientists and business analysts for Machine Learning & Statistical analysis using Python & R. We can now add Jupyter Notebook as an external tool in Power BI and access the Power BI dataset directly in notebook for data manipulation, advanced visualization, statistical analysis and machine learning models. In this session, I will show you how to add Jupyter Notebook as an external tool, and the use cases.
Power BI AutoML Model Interpretability & Explainability
Power BI AutoML allows BI developers and data analysts without significant Machine Learning experience to train & deploy machine learning models in Power BI service. Analyst's ability to debug, explain & interpret the model results is as important as model accuracy to improve the predictions and also draw key insights from the data that business users can use. In this session I will discuss ways 'Interpretability & Explainability' can be used in PowerBI AutoML.
Model Interpretability using AzureML SDK
Machine learning model interpretability and explainability is as important model accuracy. It not only helps debug the model performance but can also help draw key insights from the data that business stakeholders can use to make strategic decisions. In this lightening talk, I will show how to use the model interpretability library in AzureML and its use cases.
Mastering PowerBI AutoML
With the Premium Per User licensing, PowerBI users now access to Premium features such as AutoML which helps analysts create Machine Learning models in Power BI. Although AutoML is easy to use and setup, in this session I will show how to use it correctly and get the most out of it. I will show some of the hidden features and pitfalls to avoid.
Key Influencer Deep Dive
Key Influencer visual offers AI capabilities within Power BI without any coding and while it is easy to setup, there is a lot going on. In this session I will demystify the Key Influencer visual and help attendees set it up correctly and interpret the results accurately.
Introduction to AzureML SDK
In this session, I will provide introduction to using AzureML Python SDK for training, deploying and monitoring Machine Learning models in Azure ML service. It can be confusing for new users, so I will simplify the sdk so it's to understand and follow.
Enterprise Risk Modeling & Analysis with Power BI
While traditional BI dashboards show "what has happened", it is equally important for companies to gauge the future uncertainty and understand "what might happen" to create robust business strategy. In this session I will show how to create Monte Carlo Simulations using statistical distributions for various business scenarios so Power BI developers can help the stakeholder create Enterprise Risk models. Power BI has limited native risk modeling capabilities but with some creative modeling techniques, the BI analysts can create sophisticated risk models.