[Music] hello and welcome everyone my name is Mia and I will be your instructor for responsible AI fairness and bias mitigation in machine learning this course is a three-day course and each day will have four modules let's have a look at the learning outcomes for this course so really we're going to have two-fold learning outcomes one is going to be the theoretical understanding of first of all what is machine learning where this buyers in machine learning come from and then the second component is all about Hands-On application and really getting to practice a lot of the theoretical techniques encoding so you're really going to get practical machine learning skills and techniques we're going to train tune test and evaluate some simple machine learning models we'll be checking data and models for buyers and the key goal really here is how to identify and mitigate bias issues in machine learning as already mentioned we have four modules in total day one fundamentals of machine learning introduction to fairness and bias mitigation in machine learning model formulation and data collection and we're going to wrap up today with exploratory data analysis so really the idea is that we're going to take a typical machine learning life cycle from ideation and formulating a machine learning problem all the way to productionizing machine learning solution and we're going to have a look at every step of the life cycle and see where could buyers potentially creep in and what can we do to mitigate so that means on day two we're going to continue with the life cycle and look at data processing machine learning algorithm selection model build and evaluation we're going to need a deeper dive into fairness criteria and think more about the mathematical notation and how to formulate fairness in machine learning and then we're going to have a look at bias mitigation during pre-processing on the final day we'll continue into bias mitigation during model training bias mitigation during post-processing and then we'll look at what we can do if a model is already in production and the final module will be explainability so providing explanations for Model results to stakeholders or customers and users of the model [Music]