Steven's Interactive Models
Interactive Models - The Future of Business Intelligence Reporting
Steven has extensive experience in business intelligence and machine learning, and has developed interactive models that offer clients a unique dashboard experience. While traditional reporting methodologies are limited in their ability to support predictive, simulation, and optimization modeling, Steven's Interactive Models enable end-users to adjust model parameters and assumptions in real-time, supporting an infinite number of scenarios. These models include predictive, neural network, supervised and unsupervised machine learning, simulation, optimization, and forecasting models, and can be tailored to suit specific business requirements. Steven's approach enables organizations to extract additional value from existing models and continue leveraging them to enhance their analytical capabilities. The models are integrated into advanced dashboards, often using Tableau, Power BI, D3, Python, R-Shiny, and other cloud services.
Examples of Steven's Interactive Models
Mitigating the Spread of Communicable Diseases
Network diagrams have various practical applications. In this video, I illustrate how risk detection models can be combined with network diagrams to display the transmission of infectious diseases in the workplace.
Creating an MLB Team to Meet Multiple Objective Functions
In this video, I delve into additional applications of my MLB General Manager Dashboard. For a comprehensive understanding of the GM Dashboard, I suggest viewing Part 1 (link below) which dives deeper into the dashboard's optimization, prediction, and clustering models.
Using Player Insights to Predict Team Wins
Using optimization and machine learning models, we've developed a Tableau dashboard that maximizes the Boston Red Sox's team roster and starting lineup based on a number of team, position, and player constraints. This dashboard uses Tableau's external client integration to recalculate model output in realtime using an R-Server.
Interact with Steven's Dashboards
When possible, Steven makes his work available for others to interact with and build upon. However, it should be noted that due to limitations in connecting to Python and R servers on Tableau Public, advanced modeling functionality may not be available on shared dashboards.
Run Machine Learning Models in Realtime
With embedded model parameters, users have the ability to run an infinite number of model scenarios in real-time, while also being able to save the model output for future reference and to grant access to others.
Evaluating the Effectiveness of Student Assistance Programs
To effectively measure the outcomes of A/B tests, it's essential to establish a specific set of metrics and track them throughout the experiment. Steven was able to assist a struggling school district in monitoring the success of their student assistance programs, using these metrics to evaluate their effectiveness before scaling the programs to all schools in the district.