Data Science Terms and Techniques

Question 1

Hypothesis

Answer 1

an assumption made about the world that can be tested using data; an educated guess that needs to be validated or disproved by experiment and data

Question 2

Statistical Inference

Answer 2

a branch of statistics dedicated to drawing conclusion about the world using smaller data samples

Question 3

Confidence intervals

Answer 3

an interval estimate used to express the degree of uncertainty associated with a sample statistic

Question 4

Statistical Significance

Answer 4

an estimate of how likely that the observed event has some kind of real world importance; an estimate of how likely an event might occur randomly - the smaller the number, the more likely that the observed event has some kind of real-world importance.

Question 5

Big Data

Answer 5

a collective term used for technology to analyze large amounts of data to unearth insights, typically into human behavior and patterns

Question 6

Data Set

Answer 6

a collection of data to be analyzed

Question 7

Analytics

Answer 7

a collective term for techniques used to analyze data, mostly to draw business insights

Question 8

Algorithm

Answer 8

a well defined set of steps to solve a specific problem

Question 9

Technology Stack

Answer 9

the collective set of tools and programs used in an organization or team

Question 10

Pre-packaged distribution

Answer 10

a package that bundles all of the required python tools and libraries e.g. numpy, scipy, pandas, scikit-learn, jupyter, matplotlib, seaborn and statsmodels. In the python world, Anaconda and Canopy are popular distributions for scientific computing and data science.

Question 11

Regular Expressions

Answer 11

a technique to quickly search for or substitute complex patterns in strings

Question 12

Jupyter

Answer 12

formerly known as IPython, this tool enables data scientists to prototype code rapidly and combine it with useful documentation

Question 13

Raw data

Answer 13

data from original or secondary sources that may be unstructured or corrupted and needs more work performed on it before it can be analyzed

Question 14

Data Wrangling

Answer 14

process of taking data in its raw form and manipulating it in various ways into a useful form

Question 15

Messy or Dirty data

Answer 15

data can be messy or dirty in the sense that it might contain values that are invalid, missing, corrupted, inconsistent or non-uniform

Question 16

Storytelling

Answer 16

this highly effective art of communication is not limited to entertainment; it is a crucial skill needed to communicate answers to questions that data scientists ask

Question 17

Visualization

Answer 17

a picture is worth a thousand words, this rings true when you are trying to present data in an easy to understand manner. Data visualization is increasingly used to depict data analysis. Think about how data is visualized during election times.

Question 18

Supervised Learning

Answer 18

algorithms that create a model of the world by looking at labeled examples

Question 19

Unsupervised Learning

Answer 19

algorithms that create a model of the world using examples without labels

Question 20

Bayesian Analysis

Answer 20

algorithms based on Bayes Theorem, which makes inferences about the world by combining domain knowledge or assumptions and observed evidence

Question 21

Clustering

Answer 21

a family unsupervised learning algorithms used to automatically find groups in data sets

Question 22

What does data science entail?

Answer 22

processing, analyzing and visualizing data.

Question 23

Why is python a popular language for data science?

Answer 23

Because it can:
handle large datasets
works with common mathematical functions
creates powerful data visualizations

Question 24

What is Python Jupyter Notebook?

Answer 24

It’s built around a typical data analysis workflow and very different from an integrated development environment such as Pychar which focuses more on just working with code. In Jupyter, you work with notebooks which mix plain text, code and code outputs in one view. You can interleave code with markdown text explanations which enables you to easily explore data, create visualizations and share your results.

Question 25

What is a kernel in Jupyter Notebook?

Answer 25

The kernel defines the programming language that the code in the notebook will be written in. This is displayed in the top right corner of the notebook. When you run code, it’s executed inside the kernel session.

Question 26

What strategies should one use to address missing data?

Answer 26

1. Remove any rows that contain missing data
2. Populate the empty fields with a specified value
3. Populate the empty fields with a calculated value.
4. Use analysis techniques that work with missing data.