← Back To All Blogs

Demystify ....ing Maths for Data Science

- 2 minutes read | By Harshit Singh& Nityanand Mathur



Machine learning is built on the fundamentals of mathematics. However, you do not need to know or understand all the math to learn and code.

What do I Believe??


I strongly believe that mathematics for data science is very important but many people might find it fearsome to face the maths. It might feel that just because you don't have a solid background in maths, stats, or computer science, you can't pursue data science.

Objective / Philosophy


In this blog, I have attempted to list out basic mathematical elements you need to know before starting a data science career. Even not having expertise in them won’t affect your journey.

Content


- Algebra - It is one of the most basic domains of mathematics which is applied directly/indirectly at many places in machine learning algorithms such as calculating loss functions, regularisation, covariance matrices, Singular Value Decomposition (SVDs), Matrix Operations, and support vector machine classification, PCA to reduce the dimensionality of data. Linear Algebra is also intensively used in a neural network for processing and representation of networks.


- Calculus - In calculus, if you are familiar with calculating the derivative of a multivariate function, then it is quite easy to understand the concept of calculus in Data Science. One example is performing Gradient descent for calculating the minima of a loss function. The essential calculus required for data science is Differential and Integral Calculus, Partial Derivatives, Vector-Values Functions, and Directional Gradients.


- Probability - Uncertainty is important to machine learning. There is 3 common uncertainty in data science a) noisy data b) limited coverage of the problem area c) imperfect models. Using correct probability techniques we can easily solve the problem of uncertainty.


- Statistics - this is the part, most people find difficult. The critical part needed for machine learning include distributions, discrimination analysis, and hypothesis testing.
Fundamental statistics needed for ML are Combinatorics, Axioms, Bayes’ Theorem, Variance, Expectation, Random Variables, Conditional, and Joint Distributions.

Next step


Complete Maths Tutorial required to start Data Science and Machine Learning......Comming Soon






Comment about the blog👇

Knowledge is power. Knowledge shared is power multiplied.




 

Work done byHarshanz for iamdata