Statistical and Mathematical Foundations of Data Science and Machine Learning
Getting to know the Mathematics and Statistics behind the ML algorithms. Appreciation of these conce...
COURSE OVERVIEW
Getting to know the Mathematics and Statistics behind the ML algorithms. Appreciation of these concepts helps in 1. Understanding how the algorithms learn from data 2. Explaining how the results have been obtained 3. Selecting appropriate evaluation parameters Discussion of these topics are done in the context of solving business problems with Machine Learning algorithms and theoretical depth is determined by these considerations. While the Foundation module is adequate for practitioners, researchers in Machine Learning need to take higher level courses as well.
Linear Algebra |
How AIML problems are formulated using the concepts of Scalar, Vector, Matrices and Tensors |
Understanding how Matrix multiplication, identity, inverse and other related concepts such as norm, span, dependence and determinant accelerate AIML algorithms |
|
Probability |
Probabilistic nature of machine learning - understanding how theory of probability forms the basis of predictive analysis |
Applying probability distributions, marginal and conditional probability, Bayes' rule to solving simple problems and how it is scaled up to machine learning algorithms, Hypothesis testing |
|
Calculus and Optimisation |
Finding minima and maxima with calculus - how simple principles of calculus when combined with vast computing power can provide solution to complex problems |
Stochastic Gradient Descent Optimisation, Constrained Optimisation and Linear Least Square |
|
Statistics for Machine Learning |
Measure of Central tendency - Mean, Median, Mode |
Dispersion Measures, Range, Variance, Covariance, Standard Deviation, Z-score, Kurtosis, Skewness |
|
Correlation and covariance, VIF |
|
Normal distribution and other types of distribution, z- score and cumulative distribution function |
|
RMSE, MAE, R-squared |