Supplementary
Additional contents and supplementary materials!
One can argue that Artificial Intelligence is nothing but pure mathematics! The more math you know, the easier studying this course and any other engineering course would be for you. No significant background in mathematics is necessary for taking this course, but students are expected to be familiar with basic Calculus. In addition, knowing some Linear Algebra is a plus.
If you feel uncomfortable with (Multivariate) Calculus, these courses and materials can be helpful:
An introductory calculus course by Prof. David Jerison, covering differentiation and integration of functions of one variable, with applications.
A course by Prof. Denis Auroux that covers differential, integral and vector calculus for functions of more than one variable. These mathematical tools and methods are used extensively in the physical sciences, engineering, economics and computer graphics.
This book covers all of Undergraduate Level Calculus and is highly praised for providing complete and precise statements of theorems, using geometric reasoning in applied problems, and for offering a range of applications across the sciences.
About half of this course hinges around a general understanding of probability theory and some essential statistical tools such as MLEs and Confidence Intervals.Although most of the materials in this course are self-contained and usually do a retouch on these fundamental concepts, it is highly suggested that students who take this course, already have a strong background in both probability and statistics.
There are a number of excellent courses available online covering these topics. A skim through the material or lectures of such courses might be a perfect way to review and fill in the gaps of your probabilistic and statistical knowledge!
The following two courses from MIT OCW, are a great start:
A thorough course that covers probability theory from axioms up until basic Markov Chains and Markov Decision Processes. Instructed by the great Prof. John Tsitsiklis, auditing this course is highly suggested for those students who feel uncomfortable with general probability theory.
A complete course by Prof. Philippe Rigollet, covering topics ranging from Parametric Inference to PCA and Generalized Linear Models.
If you are looking for courses taught in Farsi, Prof. A. Sharifi's Engineering probability and statistics course on maktabkhoone is highly suggested.
There is no doubt that consulting one of the numerous well-known textbooks on each of these subjects is also a great way to ensure a concrete understanding of probability and statistics. The following is a list of suggested references to checkout:
features clear and intuitive explanations of the mathematics of probability theory, outstanding problem sets, and a variety of diverse examples and applications. This book only assumes a background in elementary calculus.
This book covers a much wider range of topics than a typical introductory text on mathematical statistics. It includes modern topics like nonparametric curve estimation, bootstrapping and classification, topics that are usually relegated to follow-up courses. The reader is assumed to know calculus and a little linear algebra.
ML In Continuous Space
ContentIn this notebook we talk about gradient descent, challenges and improvements on this algorithm.
Convex Optimization
ContentIn this notebook we talk about Convex Optimization fundamentals. We also examine a simple practical application of convex optimization; Image in-painting using CVXPY package in Python.
Prolog - Gateway to Logic Programming
ContentIn this notebook we talk about Prolog basics. We also learn how to benefit from Prolog's power using Python.
HMMs' Applications - Speach Recognition
ContentIn this notebook, we will study how Hidden Markov Models can be applied to speech recognition and introduce some useful automatic speech recognition (ASR) tools.
Machine Learning - Decision Trees
ContentIn this notebook, we will study decision trees and different types of them and we will discuss topics like pruning and indexing in ML.
Robust and Trustworthy Machine Learning
ContentIn this notebook, we will look into two of the main topics in robust and trustworthy ML, evasion and poisoning attacks, and mechanisms to defend against them.
Deep Reinforcement Learning
ContentIn this notebook, the first applications and attemps to use DL in RL are explained. Moreover, there is a practical example of DQN in this notebook.
TODO