[MUSIC] Hi, my name is Anton and welcome to the course Calculus and Optimization for Machine Learning from Higher School of Economics. Our course primarily aims to build up the necessary mathematical base for data analysis and machine learning. [INAUDIBLE] he becomes one of the most essential modern end motivations for studying calculus, the idea for optimization. In various machine learning tasks, we find ourselves in position of initial assumption of the model, the function trying to mimic the rule that produces given output from given input. For example, distinguish between cats and dogs in the picture with a set of parameters. Even come up with the idea of models, arrows and it is essential to think about the best set of parameters in order to minimize it, or in other words, trying to publish. This optimization procedure even numerical is rather complicated, and that requires extensive mathematical understanding in all kinds of line concepts. In doing so, it becomes particularly as that in our day to day life we are surrounded by impressively comprehensive function. For example, affirmations, composite arrows, prediction of some algorithm, dependable sometimes on millions of parameters. So moral price in the stock markets is dependable on various indicators of the market today and sometimes wants to interfere. Let's go over to some level as a function of sometimes missing measurements in numerous geographical locations. All this implies that we are in desperate need of the idea of proper approximations with simpler functions with the same properties as the initial complex ones. For example, finding along the approximation to define the slope or the function is instantaneous speed of change. And of course, misuse of goals by introducing consequently basic functions, concept, the idea of limits, then proceed with lightning approximations and integrals thus building up the base for all might optimization procedure. And the contradiction of this usual calculus courses grew up to embrace that in order to bring the discussed material to real life task, we need to simultaneous to discuss functions of cinco and several variables. All these include extensive non-monitored practical quizzes and interactive plots, they used in our slides for you to play with. And thus understand the concept practice in about more clearly. As a final project, we propose a Python programming task on classical optimization problem from machine learning. Hope to see you soon in the first week.