Chapter 1 - Business Analytics (BA) at a Glance

BA vs. BI

BA is more about anticipated future trends of the key performance indicators. This is about using the PAST data and models to make predictions. (different from the reports in BI)

broad umbrella BI

BA, data analytics, advanced analytics

BA addresses a number of questions

What is happening and Why did something happen? Will it happen again? What will happen if we make changes to some of the inputs? what the data is telling us that we were not able to see before?

Business Analytics (BA)

a data driven decision making approach that uses statistical and quantitative analysis, information technology, and management science, along with data mining and fact-based data to measure past business performance to guide an organization in business planning and effective decision making.

artificial neural network learning algorithm (ANN/NN)

a learning algorithm that is inspired by the structure and functional aspects of biological neural networks

machine learning

a method used to develop complex models and algorithms that are used to make predictions

classification

a process of assigning items to pre-specified classes or categories a supervised learning technique where a training set is used to find similarities in classes

machine learning tasks

classified into 3 broad categories, depending on the nature of the learning "signal" or "feedback" available to a learning system -supervised learning, unsupervised learning, reinforcement learning

clustering vs. classification

clustering- an unsupervised learning technique used to find groups or clusters of similar instances on the basis of features classification- a supervised learning technique used to find similarities in classification based on a training set

Time Series Analysis and Forecasting Models

commonly used forecasting models are regression-based models that uses regression analysis to forecast future trend

prescriptive analytics

concerned with optimal allocation of resources in an organization tools- used to optimize certain business processes and use a number of different tools that depend on specific application area -optimization models, simulation models, decision analysis, operations management (planning, analysis, control), spreadsheet models

prescriptive analytics (BA tools)

concerned with optimal allocation of resources in an organization; optimizing and automating business processes ex. linear/non-linear optimization models, operations management

deep learning

consists of multiple hidden layers in ANN; this approach tries to model the way the human brain processes light and sound into vision and hearing applications- computer vision, speech recognition

descriptive analytics

graphical and numerical methods and tools in BA; involves the use of descriptive statistics including the graphical and numerical methods to describe the data used to understand the occurrence of certain business phenomenon or outcomes and explain these outcomes through graphical, quantitative and numerical analysis tools- graphs and charts along with some newly developed graphical tools such as bullet graphs, tree maps, and data dashboards

data mining

involves (i) extracting previously unknown and potential useful knowledge or patterns from massive amount of data collected and stored, and (ii) exploring an analyzing these large quantities of data to discover meaningful pattern and transforming data into an understandable structure for further use

Arthur Samuel

machine learning gives "computers the ability to learn without being explicitly programmed" -1959 IBM

machine learning vs data mining

machine learning= used for prediction, based on *known* properties learned from the training data data mining algorithms= used for discovery of (previously) *unknown* patterns

Big Data

most businesses collect and analyze massive amounts of data using specially designed big data software and data analytics

unsupervised learning

no labels are given to the program; the learning algorithm is expected to find the structure in its input goal- finding hidden pattern in the large data

anomaly detection

outlier detection; used to identify specific events, or items, which do not conform to usual or expected pattern in the data

Data Mining

part of predictive analytics; used to analyze business data; software used to analyze vast amount of customer data to reveal hidden patterns, trends, and other customer behavior

clustering

technique is used to find natural groupings or clusters in a set of data without pre-specifying a set of categories' an unsupervised learning technique where a training set is not used

predictive analytics

the application of predictive models to predict future business outcomes and trends tools- regression models (simple, multiple, non-linear, with indicator variables), time series analysis and forecasting, simulation models, statistical inferential tools

cluster analysis

the assignment of a set of observations into subsets (called clusters) so that observations within the same cluster are similar according to some pre-specified criteria, while observations drawn from different clusters are dissimilar

supervised learning

the computer is presented with example inputs and their desired outputs, given by the analyst, and the goal is to learn a general rule that maps inputs to outputs

reinforcement learning

the designed computer program interacts with a dynamic environment in which it has a specific goal to perform -NO input/output pairs provided so there must be a balance between exploration and exploitation (current knowledge)

association learning

used to identify the items that may co-occur and the possible reasons for their co-occurence classification and clustering techniques used

predictive modeling (predictive analytics- BA tools)

used to predict future business phenomenon ex. spam detection in messages, CRM, predicting customer buying patterns, data mining

BA tools

used to visualize and explore the the patterns and trends in the data to predict future business outcomes with the help of forecasting and predictive modeling

data mining

uses many machine learning methods; concerned with knowledge discovery in databases (KDD)

predictive modeling

uses statistical models, such as, different types of regression to predict outcomes and is synonymous with the field of data mining and machine learning

descriptive analytics (BA tools)

uses statistical, graphical, and numerical methods to understand the occurrence of certain business phenomenon