Video: Introduction to Machine Learning

This 30 minutes video features my interview about the upcoming course “Intuitive Machine Learning and Explainable AI”, based on my new book with the same title. The course is described here, and the book is available here. Course participants get a free copy of the book. Both the course and the book include a solid introduction to scientific computing in Python.

Victor Chima, co-founder of, interviewing Vincent Granville

In this presentation, I answer the following questions:

  • My background: computer vision, natural language processing (keyword scoring, web crawling), computational statistics and scoring technology. Worked at Visa, Wells Fargo, NBC, eBay, Microsoft.
  • Important business problems that the course can help participants solve: taxonomy creation (NLP), fraud detection, shape recognition (computer vision), scoring using boosted / ensemble methods, leveraging rich synthetic data to get more robust predictions, automation of data validation / data cleaning / exploratory analysis, enhanced visualizations with data videos, and model-free inference techniques to name a few.
  • Why is it important for AI models to be explainable (or in some cases, legally required).
  • Exciting new trends in Machine Learning: synthetic and augmented data, generative models, automated data cleaning, causality detection, explainable AI.
  • What is unique about the course that participants can’t get elsewhere? Quickly getting to the point, very little statistics used even when discussing advanced topics. Enterprise-grade projects, understanding inner workings of Python libraries rather than using them as black-boxes, creating your own Python library. Finding simple solutions when possible, learning how to self-learn, and getting personal LinkedIn endorsement / job recommendation from a top ML influencer upon successful completion of the course. Last but not least, you will learn a new regression technique that encompasses and unifies all existing methods in one simple optimization algorithm, and a simplified yet powerful alternative to XGboost.
  • Anything else you would like to say to your future students? You will learn how to build a portfolio on GitHub, build a strong presence on LinkedIn, blend your code with third party solutions, and learn how to efficiently find the free, good quality resources to solve new problems not discussed in any book or classroom.

%d bloggers like this: