What Really Data Science Is?

### Blog > Data Science

First, let's explain what data science is not.

* Data science is not a science.

* Data science is not magic.

* Data science is not machine learning.

* Data science is not statistics.

* Data science is not software engineering.

* Data science is not programming.

* Data science is not big data.

Data science is not dependent on a specific subject area, it can be used in various fields / scenarios from cancer diagnosis to social media analysis, therefore it requires knowledge of many subjects in different fields from the fundamental to the detail depending on the problem.

Definition: Data science is a multidisciplinary science that consists of the combination of

a. Scientific problem definition and solving methods,

b. Mathematics: real and complex numbers, series and sums, logarithms, exponentials, polynomial functions and rational numbers, fundamental geometry, theorems and trigonometry, limit, derivative, integral, sets and set operations, proof and proof methods,

c. Statistics and Probability: data summarization and descriptive statistics, standard deviation, variance, covariance, correlation, sampling, measurement, error, random number generation, probability calculus, bayes theorem, conditional probability, probability distribution functions: uniform, normal, binomial, Chi-square, central limit theorem, A / B test, hypothesis test, confidence interval, p-values, linear regression, regularization,

d. Computer sciences (algorithm design, data structures, database systems, system architecture, network programming),

e. Software development.

Data science helps to transform both structured and unstructured data into useful / meaningful / valuable / insightful knowledge to solve complex problems.

Dividing a data science project into 5 different phases and proceeding step by step will enable us to abstract the problem and lead us to more effective results. These stages can be as follows.

1- Problem Definition

2- Literature and Best Practice Search

3- Data Preparation

4- Method and Algorithm Selection

5- Performance Tuning

If you have questions don’t wait: You can send us a mail at [email protected]