Although most of my practical machine learning experience stems from my personal projects, my full stack software engineering job has presented me with opportunities to create machine learning models to augment existing software. I also undertake freelance machine learning engineering and data science work that enables me to further refine my skills and immerse myself in this dynamic, fascinating world. My paid work—while increasingly focused on natural language processing—spans the entire data science-machine learning engineer spectrum, covering deliverables as disparate as neural networks, data visualization documents, Bayesian networks, regression reports, and support vector machines:
• Designed and implemented a k-nearest neighbors classifier that utilizes current river streamflow and weather data to accurately predict whether future streamflow levels will satisfy a regulatory threshold.
• Created and optimized a linear regression model to estimate demand for a regional food manufacturer's products, enabling it to fine-tune pricing and intelligently plan future expansions.
• Employed R to perform exploratory data analysis for a wholesale distributor before utilizing Tableau for data visualization, generating aesthetically pleasing, user-friendly graphs to empower business leaders and decision makers.
• Currently finalizing sentiment analysis software that utilizes a Bayesian network and support vector machine to classify user reviews, empowering a regional food manufacturer to accurately ascertain its users' preferences and quickly address concerns.
• Presently designing a recurrent neural network—based on the Long Short-Term Memory architectural paradigm—to actualize a domain- specific chatbot for a regional food manufacturer, enabling customers to receive immediate responses to their questions.