Trent Lemkus Ph.D's Linkedin Analytics

I am always eager to apply my unique combination of sales expertise and analytical skills to boost both revenue and customer satisfaction in my next role. With a strong background in Statistics and 5+ years in tech sales, I've applied machine learning and data science to consistently drive multi-million dollar revenue growth. My expertise in R, Python, and AWS SageMaker pairs seamlessly with a proven sales strategy, evidenced by notable accomplishments like reducing hedge fund losses by 5%. PhD Research: SVEM is an ensemble modeling approach developed to improve prediction accuracy and model selection for design of experiments (DoE). It works by constructing a cohort of models using various machine learning algorithms and then leveraging a self-validation process to determine the optimal weighting of models in the ensemble. This weighting is based on each model's predicted residual error rather than traditional metrics like R-squared. The ensemble model with optimized weightings minimizes overfitting and reduces variance, improving overall predictive performance particularly for smaller datasets. SVEM provides a novel ensemble learning framework that enhances model selection for DoE and increases generalizability for robust predictions.