Data Science
Artificial Intelligence (AI), Machine Learning (ML) & Deep Learning (DL)
Data Science is being used more now with large corporations to tackle challenges like managing the Company's Supply Chain. Traditionally, Data Science has been used in research, analysis, design, building, testing and maintaining products and/or services. We focus on the business-side with Enterprise Data Science to assist with integrating supply chains, industrial data and other enterprise projects to increase, or meet, quality output while managing costs. Covering Operating Plans, Marketing, Sales, Purchasing, Warehousing, Production, Distribution and Customer Service.
Creates New and Computation Metadata
Data Science working closely with Data Management for structured analytic data models. Data Science activities range from working in R and Python to SQL data models for computational statistics. For example, range from Market Basket Analysis (MBA) to Statistical Process Control (SPC).
Computational Statistics
We typically use a two-dimensional (2D) space or plane or X,Y plot to correlate or associate variables. Moving beyond two can get difficult to understand. Data Science can assist in the database with computations across many records, dimensions, dates and numeric values. Data Science is related to "Computational Statistics" which is the use of computers and statistics. Alexicon can assist with moving data science code from frontend desktop workbooks to the Enterprise Data Warehouse (EDW) and/or Data Lake (DL). The EDW is a rich data source for structured data science computations (models/schemas) and process analytics. The EDW provides a central service for analytics and data science users. One source. Important when running big backend computations across the Enterprise and when users interact with speed and accuracy.
Alexicon uses SQL, R, Python and Scala for EDWs and Enterprise Data Lakes (EDLs). Excel, W3Schools, SQL, Python and R examples are used below.
Data science is a "concept to unify statistics, data analysis, informatics, and their related methods" in order to "understand and analyze actual phenomena" with data. It uses techniques and theories drawn from many fields within the context of mathematics, statistics, computer science, information science, and domain knowledge. - Wikipedia
Simple Linear Regression
Simple Linear Regression is basic and used extensively in Data Science. Below is an Excel and Python (Anaconda Jupyter notebook IDE) example. Simple Linear Regression is used across many data science areas.
Multiple Regression
Multiple Regression is similar to linear regression and used for prediction by using more than one independent value.
AI applications include advanced web search engines, recommendation systems (used by YouTube, Amazon and Netflix), understanding human speech (such as Siri or Alexa), self-driving cars (e.g. Tesla), and competing at the highest level in strategic game systems (such as chess and Go), As machines become increasingly capable, tasks considered to require "intelligence" are often removed from the definition of AI, a phenomenon known as the AI effect. For instance, optical character recognition is frequently excluded from things considered to be AI, having become a routine technology. - Wikipedia
AI covers many areas. Our data science focus is around Analytics, Predictive Analytics, Forecasts, Root Cause Analysis and Enterprise Optimization.
Machine learning algorithms build a model based on sample data, known as "training data", in order to make predictions or decisions without being explicitly programmed (note the projection lift).
Deep learning (also known as deep structured learning) is part of a broader family of machine learning methods based on artificial neural networks with representation learning. Learning can be supervised, semi-supervised or unsupervised. This is similar to Machine Learning above with Test and Train manually setting categories or automatically creating them working with both data or images.
Designed to establish causal relationships and identify cause-and-effect relationships. While typically used in Six Sigma, DOE is a powerful data collection and analysis technique that can be used in a variety of experimental situations. It allows for multiple input factors to be manipulated, determining their effect on a desired output (response). Below an example of data used with DOE. Minitab is a popular tool for DOE along with R and Python.
Enterprise. Orchestrated.
Enterprise Operating System (EOS)
Enterprise Resource Planning (ERP) and surrounding systems: Customer Relationship Management (CRM), Supply Chain Management (SCM), Human Resources Management System (HRMS), Information Technology Service Management (ITSM), other COTS and In-house Built Applications.
Integrated Applications and Analytics.
In-house and Cloud Planning, Migrations and System Operations.
Process focused Work Instructions.
Leverage a Data Dictionary and "Metadata" for the Enterprise to assist with planning software applications and the data landscape.
Enterprise Metadata >
Process
Enterprise Business Processes and Technical Processes for Systems and Data.
Align key processes across the Enterprise to achieve optimal performance. Sync Enterprise Applications and Analytics.
Time intelligence speeds operations and provides clear communication with Associates, Customers, Partners and Suppliers.
Formal Lean Six Sigma or streamlined methods and techniques.
Define, Measure, Analyze, Improve and Control
Statistical Process Control (SPC): Run Charts, Control Charts and Design of Experiments
Gain the process advantage and include learning in the EDW and Data Lake.
Analytics
Enterprise Visualizations, Dashboards and Reports.
Data Science
Leverage Desktop Excel, SQL, R, Python and Scala code. Hyperscale activities.
Integrate Data Science activities in the EDW for Enterprise-wide computations.
Data Mgmt.
Enterprise Applications, Other Data, EDW and Data Lake.
Data Roadmap >
