MachineIntelligence powerpoint presentation

lakshmana@teloxis.com
Use and Implementation of
Computational Intelligence
S. Lakshmanaraj
9225518035

Use and Implementation of Computational Intelligence
3/2/2024 S. Lakshmanaraj 2
• Agenda
• Components of Computation Intelligence
• Computation Intelligence-Data Analytics
Maturity Path
• Example Applications of Computation
Intelligence
• IoT
• Healthcare
• eCommerce
• Finance
• Cyber Security
• Education
• …and so on
• Architecture to Implement
• Business Component Architecture of Computation
Intelligence
• Connectivity Architecture of Computation Intelligence
• Data Horizons
• Implementation Approach
• 9 Steps Approach

Computational Intelligence
• Fuzzy Logics
• Approximate reasoning and Decision
making
• Neural Networks
• Data analysis, Classification, Associative
memory, Clustering generation of
patterns and Control of patterns
• Evolutionary Computation
• Natural evolution to bring up new
artificial evolutionary methodologies
• Learning Theory
• Process of bringing together
behaviorism, cognitivism, constructivism
along with emotional and environmental
effects
• Probabilistic Methods
• Randomness to predict the problem and
prescribe the solution combining
mathematical relations and or above
methods
STEP 1
Data
Informatio
n
Knowledg
e
Units
Collection
s
Relations
Patterns
Skill
Applications
Dimension
STEP 2
STEP 3
STEP 4
STEP 5
Connectedness Understanding
Wisdom
Principles
Decisions
STEP 6

High Level Machine Intelligence - Data Analytics Maturity Path
STEP 1
Descriptive
Analytics
Diagnostic
Analytics
Predictive
Analytics
Prescriptive
Analytics
Feeling and
Thinking
What
Happened?
Why Did It
Happen?
What Will
Happen?
How Can We
Make It
Happen?
Association
Analytics
STEP 2
STEP 3
STEP 4
STEP 5
Value
Difficulty

Where Can be Used for IoT?
• Information – Diagnostic Analytics
• Moving Speed Detections as well as oscillation frequencies
• Removal of Data noise and Self Correctness
• Growth/Decline rate – Support Cases, Manufacturing Defects rate,
Devices Wear & Tear Rate, Financial Growth
• Knowledge – Activation Functions for AI/ML
• Preventative Maintenance Schedule - modelling by sound and
temperature in motors of fan, washing machine, fridge etc.
• Prescriptive Methods - Auto switch on/off A/C based on temperature,
products pair well together and how to price products

Where Can be Used for Healthcare?
• Clinical Document Quality Index
• Growth/Decline rate – Support Cases, Recovery rate, Readmission rate,
Financial Growth
• Preventative and Corrective actions – Diagnosis data with Patient
education materials
• Predictive Methods- Number of patients visiting hospitals, Diseases
seasonal patterns
• Prescribing Methods – Number of resources needed like beds, pills,
injections, nurses etc.

Where Can be Used for e-Commerce?
• Optimal Logistics Route planner
• Decoration Pattern to connect irregular shapes
• Product Grouping to maximize Buyers and to minimize stock
• Growth/Decline rate – After sales support cases, Financial Growth
• Predictive Method- Where to invest money, Which products can be
retired, Customer segmentations
• Prescribing Methods – Price response functions, Supply and Demand
generating seasonal patterns

Where Can be Used for Cyber Security?
• Network (network traffic analysis and intrusion detection)
• Endpoint (anti-malware)
• Application, Users, Process (anti-fraud)
• At Rest, At Transit or Historical
• Prediction Methods – Anomalies, Forensic analysis
• Prescribing Methods – Encrypted Blockchain

Where Can be Used for Education?
• Digital Library
• Questions, Answers
• Markings / Categorization as Easy to Difficult from Novice to Expertise
• Prediction Methods – Most wanted materials, Attendance, Productive
hours, teaching preferences
• Prescribing Methods – Assigning Education Materials to overcome
Weak Skills, Auto scheduler

Where Can be Used …and So on…
Society
1.0
•
Hunter-Gatherer
Society
Society
2.0
•
Agrarian
Society
Society
3.0
•
Industrial
Society
Society
4.0
•
Digital
Society
Society 5.0
•Machine
Intelligence
to expand
human
capabilities
and
address
social
challenges

High Level Example of Preventative and Maintenance System
Cloud SQL
Server
Sensitive
Pump Data
Predictions/
Other data
Trained
ML
Data
Streaming
Anomaly
Detection
Real-time
actions
Maintenance
Shut-off
Core Engine
AI built-in
Machine
learning for
low-latency
analytics
Built-In Time-Series
Streaming and
Analytics
Cloud
SQL
Time-
Series
Time-series
built-in
Unparalleled
performance and
security
Most secure with
industry leading
performance

High Level Generic Business Components Architecture
DATA SOURCES
OBJECTS STORAGE
RDBMS DATA WAREHOUSE
NoSQL
COLLECT
(Ingest)
COLLECT
(Store or
Destroy)
ORGANIZE
(Transform & Manage)
FUNCTION AS
A SERVICE
WORKBENCH
DATA PREPERATION
BIG DATA
AS A SERVICE KNOWLEDGE
CATALOG
EVENT STREAMING SERVERLESS ETL
LOG ANALYSIS IoT DATA
EXTERNAL DATA
KEY
MANAGEMENT
AUDITING
IDENTITY & ACCESS
MANAGEMENT
AI RULE ENGINE MACHINE LEARNING
ANALYZE & INFUSE
(Present & Interact)
AUTOMATE PROTECT
TRUST MONITORING
RULE EDITOR
TIME SERIES DATA
RELATIONAL DATA
UNSTRUCTURED DATA
ACTIONABLE DATA
MEANINGFUL DATA
OBJECTS AT REST
X

High Level Generic Connectivity Architecture

Data Horizons
Data
Availability
Data Quality
Data
Consistency
Data Security
Data
Auditability
Reporting, Analytics and Data Science Services
Data Lineage
Data
Standards
Data Policies
and Procedures
Business
Metadata
Technical
Metadata
Master Data Management, Metadata Management and Data Governance
Reporting Self-service BI Open Data API Data Science
Relational Dimensional In-memory Polyglot
Data Architecture and Data Technology
Structured Data Unstructured Data Semi-structured Data Binary Data
Data Integration, Data Services and Data Adapters
Internal External Third party Future M&A
Data Sources
Data Grouping
and Indexing
Data Sharing
Process

High Level Generic Data Analytics 9 Steps Approach
1. Identify the problem and the stakeholders
2. Identify what data are needed and where
those data are located
3. Develop a plan for analysis and a plan for
offline or periodic or near-time or real-time
data retrievals or access
4. Extract, transform, load the data
5. Check, clean and prepare the data for analysis
and automate in minimizing time
6. Analyze and interpret the data
7. Visualize the data
8. Disseminate the new knowledge
9. Implement the knowledge in the organization

1st and 2nd Step Data Points for UN SDG Goals

Data Analytics Initial 1st to 5th Steps - IoT

Data Analytics Initial 1st to 5th Steps - Healthcare
Clinical Intelligence
Data access
Data/text search & mining
Data/text analysis
Access & Collaborative Portal
Data
Integration
Data
Access
Clinical Data
Public data, disease, procedure,
provider, treatment plan, lab
tests, patient history, etc.
Data
Acquisition
Payer
Intelligence &
Decision
Support
Financial Data
Code, Operation, Billing, Fee,
Coverage, Risk Factor, Competitor,
Customer Background, etc.
Clinical Decision Support
Evidences & Knowledge
-
Guidelines
-
Rules
Integrated Clinical Data Warehouse
(HIW, CG, HCN etc.)
Physicians/Admin Patients
Analysts and Reviewers
Patients
Solution
Portfolio
CLINICAL
Informatics
RESEARCH
Informatics
ADMINISTRATIVE
Informatics
Results, medications
Vitals, Allergies
DC Summaries
Clinical Doc
Proteomics,
SNPs,
Publications
Clinical Trials
PubMEd
ADT,
Demographics
Provider,
Scheduling
Diagnosis, CPT,
AR, Billing,
Claims
Health
Analytics

Data Analytics Algorithms Used in Step 6
Supervised Unsupervised
Linear Nonlinear
Single Combined
Easy to Interpret Hard to Interpret
Linear
Regression
Logistic
Regression
Perceptron
Bagging Random Forests Boosting
(Bootstrap
Aggregating)
Decision Trees Rule Learning Naïve K-Nearest
Bayes Neighbors
Multi-Layer SVM Perceptron Hidden Markov
Perceptron
K-Means Expectation Self-Organizing
Maximization Maps

Data Analytics Mid 6th and 7th Steps
Marts/Cubes
Reports
Portal/Kiosks
Biz Apps
Dashboards
EDW Unstructured
Big Data
Integrated Analytics

AI/ML Initiatives for Step 7
01
02
03
04
08
09
10
11
05 13
06 14
07 15
12
16 20
17 19
18
Unsupervised
learning
Reinforcement
Learning
Supervised
learning
Dimensionally
Reduction
Clustering Regression
Classification
01. Feature Elicitation
02. Structure Discovery
03. Meaningful compression
04. Big data Visualisation
08. Objects Classification
09. Objectives fulfilment
10. Diagnostics
11. Fraud Detection
19. Skill Acquisition
20. Robot Navigation
05. Recommended Systems
06. Targeted Systems
07. Segmented Systems Psychological
models
data
mining
Cognitive
science
Decision
theory
Information
theory
databases
Machine
Learning
neuroscience
statistics
evolutionary
models
Control
theory
12. Forecasting
13. Predictions
14. Process Optimization
15. New Insights
16. Real-Time Decisions
17. Gamifications
18. Learning Tasks

Data Analytics Final 8th and 9th Steps
• Disseminating
the new
knowledge
• Write up the
findings
• Disseminate to
the
stakeholders
• Implementing
the new
knowledge
• Requires
participation
of
stakeholders
Targeted
Content
Compelling Calls
to Action
Predictive
Analytics
Precision
Marketing
Findability
Self-
Management
Tools
Collaborative
Interactions
Behavioral
Insights
Advanced
Personalization

Thank
you
For more information, my Concept AI/ML activation models are published in
https://www.ijmttjournal.org/Volume-66/Issue-11/IJMTT-V66I11P502.pdf
You can connect me at
https://www.linkedin.com/in/lakshmanarajsankaralingam/
Final Thoughts... Any Questions?

MachineIntelligence powerpoint presentation

Recommended

Recommended

More Related Content

Similar to MachineIntelligence powerpoint presentation

Similar to MachineIntelligence powerpoint presentation (20)

Recently uploaded

Recently uploaded (20)

MachineIntelligence powerpoint presentation