BI, Data Analytics, Predictive Analytics, Data Mining, Machine Learning and Big Data
It is indisputable that decision-making requieres to have data and analyze this data. Excel has its limits, misinformed intuition is not reliable or infallible, and we all have plenty of data sources available today.
Data analysis has evolved with computing power, internet andcloud; new analysis tools have emerged while a whole series of techniques and terminology related specializations were born in the last few years.
The terms that define the various activities in this filed are still evolfing and sometimes they overlap and the frontiers between them are confusing.
Business Intelligenca, Data Analytics, Data Science, Data Mining, and Big Data are often used indiscriminately and sometimes we find difficulties to define or categorize a particular activity.
It makes sense to classify the terms and to analyze the scope of each of these activities.
The term Business Intelligence is old and we can be found the first references in 1860 , although the term as we understand it today comes from the use made by Howard Dresner, to describe “concepts and methods to improve business decision – making using systems that are support on facts “.
The term BI is confused, used or exchanged for Business Analytics and Data Analytics.
BI is a technique strongly supported by technology tools to analyze data and present information for users to take decisions based on measurable results and verifiable facts.
The main BI tools are databases, data queries, KPIs, reports, dashboards and visualizations.
The potential benefits of BI include acceleration in decision-making, optimizing internal processes, identifying trends and troubleshooting problems or inefficiencies.
Through BI techniques you can answer the following questions:
What happened or is happening? Who did it? How many? Where? How?
Data Analytics, DA, is a modern term that includes activities related to data analysis in order to draw conclusions but especially for analyzing causes and predict future behavior.
The questions we try to answer with Data Analytics are:
Why it happened? It will happen again? What if we change x? What else can be inferred from the data we have but we did not ask?
DA focuses on inference, the process of drawing conclusions based on what is known.
There are several stages or identifiable different activities DA
EDA, Exploratory Data Analysis, which is trying to find new data features.
CDA, Confirmatory Data Analysis, where assumptions are checked EDA.
QDA, Qualitative Data Analysis, which sets no numerical data as text or multimedia are analyzed.
Predictive Analytics, PA, is a specialized branch of the DA that focuses on making predictions about future behavior.
PA uses many different techniques such as Data Mining, Statistics, Modeling, Machine Learning and Artificial Intelligence to analyze current and past data to make predictions.
Data mining is the science that searches large volumes of data to find hidden patterns and relationships.
The main objective is to extract information, organized and enriched so that it can be used in other analysis activities, such as BI, Data Analytics and Predictive Analytics
Machine Learning, ML, is a particular type of artificial intelligence that allows computers to learn without being specifically programmed. The fundamental basis of ML is the data from which the algorithms can learn, act, suggest results and reprogram itself as new information arrives.
The process is similar to Data Mining in terms of working with large volumes of data to find relationships and patterns. The difference is that Data Mining is aimed at people looking for relationships and patterns, whereas the ML is a technique where a program itself discovers and exploits the relationships and patterns in data to make predictions an evolve.
Big Data is a set of techniques for working with very large volumes of data, terabytes or even larger, which can not be treated or analyzed by hand, but contain information important for data mining or other activities like Data Analysis.
Big Data is characterized by the concept of 3V, big Volume, high Velocity rate and great Variability in data type.
Time, resources and cost involved in dealing with these data volumes make them unsuitable for conventional databases and must be treated with special distributed systems, such as Hadoop, and in particular with Cloud tools.
 Miller Devens, Richard. Cyclopaedia of Commercial and Business Anecdotes; Comprising Interesting Reminiscences and Facts, Remarkable Traits and Humors of Merchants, Traders, Bankers Etc. in All Ages and Countries. D. Appleton and company. p. 210.
 Power, D. J. “A Brief History of Decision Support Systems”.