Anant Khurana — March 31, 2021
This article was published as a part of the Data Science Blogathon.

It is the mark of truly intelligent person to be moved by Statistics.

The most important aspect of any Data Science approach is how the information is processed. When we talk about developing insights out of data it is basically digging out the possibilities. Those possibilities in Data Science are known as Statistical Analysis.

Most of us wonder how can data in the form of text, images, videos, and other highly unstructured formats get easily processed by Machine Learning models. But, the truth is we actually convert that data into a numerical form which is not exactly our data but the numerical equivalent of it. So, this brings us to the very important aspect of Data Science.

With data in numerical format, it provides us with infinite possibilities to understand the information out it. Statistics acts as a pathway to understand your data and process that for successful results. Not only the power of statistics is limited to understanding the data it also provides methods to measure the success of our insights, getting different approaches for the same problem, getting the right mathematical approach for your data.

## Importance of Statistics for Data Science

Most Data Scientists always invest more in pre-processing of data. This requires a good understanding of statistics. There are few general steps that always need to be performed to process any data.

1. Identify the importance of features by using various statistical tests.
2. Finding the relationship between features to eliminate the possibility of duplicate features.
3. Converting the features into the required format.
4. Normalizing and scaling the data. This step also involves the identification of the distribution of data and the nature of data.
5. Taking the data for further processing by using required adjustments in the data.
6. After processing the data identify the right mathematical approach/model.
7. Once the results are obtained the results are verified on the different accuracy measurement scales.

The data processing from the beginning to the end of the complete cycle there is a requirement of statistics at every single step. That’s why a good statistician can be a good Data Scientist as well.

## Guide to Learn Statistics

It is always required to understand every fundamental of Statistics. However, most people are not very clear about where to start.

These are the few key concepts that are required to pace up and understanding the fundamentals of Statistics for Data Science:

#### Probability

Probability is the basic need for understanding the possibilities. To start with let us take a very basic example – What are the chances that Team A is going to win the football match against Team B. To derive this answer we might require 100 people to give their respective votes – Number of Samples. Based on those votes we can have a chance of which team can win the game.

But, in this example, we come across another very important concept which is known as sampling – identifying the right set of people to vote for the results. So, the probability is the chance of whether the event will occur or not. Depending upon the scenario we can build different solutions around this.

#### Sampling

The sampling as we discussed in the above example is identifying the right set of people. The question is what are the right set of people. Let’s continue our previous example for the above scenario we need those 100 people who have a good knowledge of Football, who knows about the history of Team A and B, Who should not be biased towards a team because of their personal preference. So, identifying the right sample can be done by various statistical approaches. There are various types of sampling methods – Simple random sampling, Systematic sampling, Stratified sampling, Clustered sampling, etc.

#### Tendency and Distribution of Data

The distribution of data is a very important aspect. The famous distribution like Normal Distribution is very significant. For example, when we talk about the distribution of height and weight of the world it is normally distributed data showing the symmetry of nature. The normal distribution has to Mean, Mode, and Median all coincide at the central peak. These data are supposed to be very accurate data. So identifying the distribution and skewness of data is a very important concept.

#### Hypotheses Testing

If we know whether to perform some action or not. Will those actions give a positive result or a negative result then we can have an added advantage of doing the right things. Hypotheses Testing gives identifying the situation where the action should be taken or not based on what results will it produce. There are other tests as well like A/B Testing, Z Test, T-Test, Null Hypothesis with similar relevance.

#### Variations

When we talk about different variations in data. We talk about the distortion, error, shift in data. Along, with the variations in data the range of the data, relationship within the data. All these accounts for the variability of data. Some of the key terms to understand here are – Variance, Range, Standard Deviation, Error Deviation, Covariance, Correlation, Causality, etc.

#### Regression

The regression in simple terms is finding out a relationship between the independent and dependent variables. Regression can be of two types broadly  – Linear Regression, Multi Linear Regression.

Linear Regression – Y = aX + C

Multi Linear Regression – Y = aX + bX1 + cX2 + …. + C

Statistics is a wide concept limiting not just to what exists but what can be derived out of existing techniques to build something new. Hence, Statistics is very important for Data Science as it helps to understand existing solutions as well as digging out new developments.

There is always a way to do it better – find it and become and innovator

The media shown in this article are not owned by Analytics Vidhya and is used at the Author’s discretion. 