Scribd
Upload
21 views14 pages

Electives Notes

The document discusses predictive analytics and data modeling. It begins by explaining predictive analysis aims to determine trends, correlations, causation or probabilities by describing what will happen if certain conditions are met. A data analyst uses predictive modeling and the five basic types are described. Later, it discusses how descriptive analytics works by tracking metrics, collecting, preparing and analyzing data to present findings. Descriptive and predictive analytics are contrasted.

Uploaded by

constantinotv5
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
21 views14 pages

Electives Notes

The document discusses predictive analytics and data modeling. It begins by explaining predictive analysis aims to determine trends, correlations, causation or probabilities by describing what will happen if certain conditions are met. A data analyst uses predictive modeling and the five basic types are described. Later, it discusses how descriptive analytics works by tracking metrics, collecting, preparing and analyzing data to present findings. Descriptive and predictive analytics are contrasted.

Uploaded by

constantinotv5
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

PREDICTIVE

Data Analytics Pipeline


1. Ask Questions
- Where do we need to bring the business in the future?
- Which business do we need to focus on?
2. Analyze the Data
- Collecting the necessary data/info to answer the question or addressing the
concern
3. Communicate the Results
- Relaying the information to the higher ups of the business which will lead to
business decisions that should benefit the business

How to ask questions?


- Descriptive: What Happened?
- Diagnostic: Why did it Happen?
- Predictive: What Happens Next?
- Prescriptive: What should I do?

The more complex an analysis is, the more value it brings.

Predictive Analysis
- The goal is determining a trend, correlation, causation or probability.
- Describes what will happen if these conditions are met
- “Could” >> “Will”, there will always be room for error
- A Data analyst uses predictive modeling.
Five Basic Predictive Analytics Model

Insights are aha moments - the revelations that open your eyes to a new way of seeing.

DESCRIPTIVE

How Descriptive Analytics Works


1. Track Business Metrics
2. Collect the data
3. Prepare the data
4. Analyze the data
5. Present the data

How do you come up with recommendations?


- Personal experience
● Our experiences might not be applicable for others.
● Memories can mislead
- Other’s experience
● Untested experiences
● Not sure if true or not
- Research
● Too much information / which one to follow
● Fake Information

= YOU WANT TO SEE EVIDENCE


Evidence (data) —> Informed Decisions —> Better Results

LEARN TO SPEAK IN NUMBERS

- Analytics is the art of asking questions and answering with data.

Data Analytics Pipeline


- Ask Questions
- Analyze the Data
- Communicate the Results

● Human element in every step


● Iterative process

1. Ask Questions
- What are we trying to solve
- What do you need to know
- Usually comes from the client / user / audience
- Requires domain expertise / experience
- “WHAT” before the “HOW” or clarity of purpose
2. Analyze the Data
- Do you have the data
- Is it reliable
- Is it clean
- Is it enough to solve the problem
- “Garbage in, Garbage out”
3. Communicate the Results
- How will you report your findings to the decision makers?
- Relaying what you found out from the information
- Ensuring it is understandable
- How will the analysis be presented and who are the target audience

Dealing with Questions


- Categorizing —> Prioritizing

Domain Research Best Practices


- Narrow your research topic before logging on
- Use a hypothesis-driven approach
- Use trusted websites

Data Analysis
General Question —> Specific Question —> Another specific question
PREDICTIVE

4 Pillars of Data Science

Data Mining
- Is the process of finding anomalies, patterns, and correlations within large data sets to
predict outcomes (SAS, 2015).
- The term "data mining" is often a misunderstood concept, because the goal is the
extraction of patterns and knowledge from large amounts of data, not the extraction
(mining) of data itself.
What is Data Mining?
- Iphone location and facial recognition of photos you take
- Book recommendations in Amazon

Cross-Industry Standard Process for Data Mining


- The CRISP-DM methodology provides a structured approach to planning a data mining
project.
Some examples:
• HR dataset with 10,000 rows/ data entry with 75 variables to be assessed
• Data: Binary Data set – yes or no questions (variables are in 0 and 1)
• Objective/ Business Question: Determine which of the variables would make a job/project
problematic?
DESCRIPTIVE

Data Visualization
- Present data
- Summarize the data
- Explore the data

Exploratory Data Analysis


- Performing initial investigations on data, in order to
● See pattern and trends
● Spot anomalies / outliers

Common tools of EDA


- Excel, Tableau, Power BI

Framework
- Load data, Choose data, Create a graph

Parts of a visual
- Title, y-axis, x-axis
Why statistics?
- Giving Opinions (your point of view) SUBJECTIVE
- What the data is saying. Dependent on your data and methodology. OBJECTIVE

Descriptive statistics
- The process of describing and summarizing the data in a dataset.

Inferential statistics
- The process of inferring information about a population from sample data.
- Conclusions.

Predictive statistics and Regression


- The process of analyzing long-term trends from historical data.
Measures of Dispersion
- A measure of how scattered or spread out the data is.
SD and Variance
- What is the average distance of our data points from the mean?
- Square root of variance

You might also like