Big data analytics by Kim H.Pries & Robert dunnigian

BIG DATA ANALYTICS:

BIG-DATA-ANALYTICS 

A PHYSICAL GUIDE FOR MANAGER

The book-big-data analytics is written by (Kim H.Pries & Robert dunnigian) 

• VARIANT 1 : download english book 

Syncing File...

Confirm Download

VARIANT 1 (English) BIG DATA ANALYTICS
Kim H. Pries & Robert Dunnigan

join our WhatsApp channel to stay updated and get every book

JOIN OUR COMMUNITY

1. The Paradigm Shifts: From "Sampling" to "The Whole"

One of the most interesting things that Pries and  Dunnigan point out early on is the change in statistical philosophy.When the “Small Data” era, we were sampling, taking a small piece of population data and mathematically estimating what the whole was enough

The authors state that  Big Data makes unnecessary most of the work of sampling. “When you have the processing power to analyze every transaction, every sensor ping, and every customer click, the margin for error shrinks, but the ‘noise’ increases., The problem they pose isn’t finding the data; it’s staying afloat in the flood without losing focus.” and that's what needed most

2. The Technical Ecosystem (Simplified for Strategy)

The book almost spends a lot of time explaining the topic like "Demystifying or explaining the Big Data Stack." they don't want managers to be programmers but they are ;adamant that a manager who does not /know the difference between organized and unorganized data is a great liability.

Structured Data: This is a  typical Excel spreadsheet or(SQL)  database. it is clean organized and easy to use also its easy to implement formulas on it 

Unstructured Data: this is the unorganized and unstructured data like a (wild stack) including mush type of data in different verities 

The Hadoop Revolution: The authors describe Hadoop as an ecosystem, not a tool in the system. They describe MapReduce (the process of taking a large problem and breaking it down into tiny pieces and solving them at the same time and then combining them) and HDFS.

The lesson for the manager is scalability. In the old days, if you ran out of space, you bought a bigger, more expensive "super-server." In the Big Data world, you simply add another cheap, "commodity" PC to the cluster. This changed the economics of business intelligence forever.

3. The "Seven Pillars" of Big Data Implementation

Pries and Dunnigan have outlined a roadmap for organizations attempting to transition from traditional reporting to advanced analytics. They propose a structured approach:

Inventory: What data do we actually have? most companies don't pay attention to there data 

Cleanliness: data can be like a water if source is polluted then the results can me distructive

Storage: Whether t-o use in-house /servers or cloud storage.

Analysis: What is the appropriate "/lens" to use are we seeking trends?

Output: converting of data in to a meaningful format is only beneficial when it serve benefit to the organization other wise its just a waste of money time and resources

4. Statistical Methods and Machine Learning

The authors dive into how Big Data changes the "Math." They discuss several key techniques in a way that’s to non-mathematicians:

• Regression Analysis: Predicting a value based on other variables (e.g., how much will weather affect sales?).

• Clustering: Finding "natural" groups in data. This is huge for marketing; instead of generic demographics (like "Men aged 18-35"), cluster is a group of selected things or persons used for for sampling 

• Association Rules: The classic "Beer and Diapers" example identifying that the things that are mostly bought together to optimize store themes or website recommendations can cause a long term benifits

5. The "People" Problem: Roles and Culture

A significant portion of the book is dedicated to the topic of the human element. Pries and Dunnigan warn that "The best algorithm in the world cannot save a company with a closed mind." so that means we have to diversify our thinking pattern to increase our knowledge and thinking 

They define the necessary roles:

• The Data Scientist: The "wizard" who knows the math and the code and can do both of them.

• The Data Architect: The person who builds the "pipes" that move the data and creates the system.

• The Savvy Manager: The bridge. This person knows enough about the data to ask the right questions and enough about the business to turn an insight into a profit for a long run.

They also tackle "Analysis Paralysis." Managers often wait for 100% certainty from the data, but the authors argue that in the Big Data world, 80% certainty with high velocity is often better than 100% certainty when the opportunity has already passed so that depends on a chance.

6. Ethics, Privacy, and Security

Written with a prophetic edge, -the authors discuss the here "creepy factor." They caution that just because a company can correlate a customer’s location with their health records and purchase history, doesn't mean they should do that

They highlight the massive risk of "Data Breaches." In the Big Data world, a breach isn't just a few leaked names, it’s a categorical exposure of a person's life. They advocate for Security by Design—building the locks while you build the database, not as an afterthought it is a task.

7. The Business Value of "V"

The book constantly and much circles back to the ROI and RI (Return on Investment) (Reseidual income). They break down the 3 Vs (plus a few extra) into business terms:

• Volume: Lowering storage costs while increasing "memory."

• Velocity: Making decisions in real-time (e.g., fraud detection on a credit card happens in milliseconds).

• Variety: Combining data from different departments (Sales + HR + Logistics) to see the "full picture"

• Veracity: How much can we trust this data? This is the authors' favorite "V," as it deals with the quality of the insights.