This talk explores the basics of AI and machine learning from an application point of view. We run through basic definitions and examples. Then we talk about management of AI/ML projects.
1. How to Boost Your Career
With AI
How deep learning and machine learning
can help you get ahead in your career
Oxygen AI
2. Goals
I. Understand what AI is and the current capability of AI
II. Be able to Assess how AI can help your work
III. Know the steps to take to Use AI
IV. Understand Risks/Caveats of using AI
3. About Me
15+ years in high tech
Currently: CEO at Oxygen AI
•BS, Physics – Florida A&M Univ.
•MS/PhD, Mechanical Engineering – Univ. of Md
•MBA – Cornell Univ.
Loves: Travel, Astronomy, Running, IoT, & Jazz
Email: keita@oxygen-ai.com
Twitter: @oxygenai, @keitabr
4. I. AI Basics & Capability
• Intro to AI, Machine Learning, Deep Learning
• How AI industry is different
• Applications and Capabilities of AI
• Images & Video
• Natural Language
• Tabular Data
• Time Series
5. What is AI & Machine Learning ?
The science of making machines replicate human
intelligence.
Artificial
Intelligence
A pillar of AI, where algorithms allow machines to
learn from data.
Machine Learning
One powerful subset of Machine Learning. An
Extended Neural Network.
A Neural Network is an algorithm inspired by the
human brain.
Deep Learning
Heavy SW Development Light SW Development No Coding
How is it
Done?
What AI is not:
• Not automation per se
• Not a Set of Rules
- Ground-up development
- Using coding languages
(such as scala, python) and
databases (SQL, no-SQL)
- Using PaaS
- Takes advantage of existing
packages and platforms (IBM
Watson/Bluemix, AWS)
- Vendor Products
- User friendly products that
require various levels of
configuration and input data
6. Machine Learning: Industry &
Methods
How Machine Learning in Industry is Different
• Open Sharing of Algorithms and Techniques
• Non-proprietary SW Platforms
• Inexpensive to Get World Class Results
Overview of Methodology
• Output: Interpretation vs. High Accuracy Predictions
• Predictions: Regression vs. Classification
• Human Guidance: Supervised vs. Non-Supervised
13. II. Using AI in your Work
• What Drives an AI Project’s Success?
• 3 Career Roles
• 3 ways to Access AI capability
• Steps to take
14. What Drives an AI Project’s Success?
• Compute Power
• Data (Collection)
• Algorithms/Code
• Integration with Actual Operations/Decision Making
Credit: Google Launchpad
15. Ways to Access AI
Develop Code Using Existing
Libraries/Modules
Cloud Platforms that can accept
light or no coding
• End-to-End Model Development: e.g.,
IBM Cloud/Watson
• Appliances for narrow applications:
Google Translate API, salesforce, etc
Hire
• Direct Employees
• Consulting Firms
• e.g., Oxygen AI
• Freelancers
17. Individual Contributor
• Goal - Enhance performance on Job
• Strategy - Leverage Domain Expertise in a Project!
• Start Small: Answer simpler questions/Use Datasets with
Fewer Features/Use simpler models
Set Core Goals Collect Data
Build Simple
Model
Non-Coder
•AIaaS Platform - IBM
Watson
Coder:
•Python - SK Learn
Modules
Evaluate Performance/
Refine Model
If Successful
Scale Up Critical
Usage of Model
18. Executive/Decision Maker
• Goal: Enhance Department/Business Unit Performance
• Strategy: Proof of Concept aligned to Org’s Goals
AssessResources
- Budget
- IT Resources
- Staff Capabilities
ProofofConcept
- Critical Mass of data
- Clear Target Output
- Model that Aligns
with business goals
- Clear Performance
metrics
Deployment
- Where does data sit
- Size of input/output
data (1 MB ; 1 TB)
- Frequency of Output
(1 per sec; 1 per month)
- How are model
outputs accessed
20. Security and Privacy
Personal Data Protection and Privacy
• Understand Regulations
• Understand the Rights to data subjects
• Make Agreements with Users
• Don’t Directly use Personally Identifiable
Information (i.e., names, email addresses,
phone numbers, etc)
• Industry Specific Regulation (i.e., HIPAA)
General Data Protection and Security
• Use of firewalls, VPNs, password manager, etc
21. Ethics in AI
• Complicated and Still Developing
• Questions to Consider:
• Bias in Data
• Bias in subjective targets
• Error rates for different sub
groups
• Processes to deal with
mistakes
Nov 2017:
Oct 2016:
Sept 2017:
Examples credit: fast.ai
22. In Summary
• AI and Deep Learning is becoming more accessible to the layman
• Domain Expertise and Data Are Key
• A good model is an essential but small part of a successful AI project
24. Life Sciences
Identifying
biomarkers
Drug/chemical
discovery
Analyzing study
results
Identifying negative
responses
Diagnostic test
development
Diagnostic targeting
Predicting drug
demand
Prescription
adherence
Putative safety
signals
Social media
marketing
Image analysis
Clinical trial design
COGS optimization
Insurance
Claims prediction
Claims handling
Price sensitivity
Investments
Agent & branch
performance
DM, product mix
Hospitality
Dynamic pricing
Promos/upgrades/o
ffers
Table management
& reservations
Workforce
management
Manufacturing
Failure analysis
Quality
management
Inventory
management
Warranty/pricing
Direct
Marketing
Response rates
Segmentations for
mailings
Reactivation
likelihood
RFM
Discount targeting
Phone marketing
Email Marketing
Construction
Contractor
performance
Design issue
prediction
Agriculture
Yield management
Automation
Mall Operators
Tenant capacity to
pay
Tenant selection
Education
Automated essay
scoring
Dynamic courses
Utilities
Optimize
Distribution
Network
Predict Commodity
Requirements
Credit: Fast.ai
25. Data Scientist
• Goal - Career; Technical Startup Founder
• Strategy: Practice, Study, Apply Knowledge, Share!
• Not overnight; but persistence will get you there!
Math
• Probability
• Statistics
• Combinatorics
• Linear Algebra
• Diff. Calculus
Coding
• Data Structures
• Functions/Classes
• Working with
SQL/noSQL DB
• Light Front End
Dev (flask; Django)
• Test Driven Dev
• Version Control
(git)
Data Science
• Cleaning Data
• Machine Learning
Algorithms
• Deep Learning
Algorithms
• Data Pipeline
Standards
• Read Relevant
Journal Articles
• Using Cloud
Platforms (AWS,
Databricks, Google
Cloud, etc)
Share
• Network (Meetups,
study groups,
company events)
• Speak at Tech
conferences
• Github Repository
• Blog about your
work (Medium)
• Follow and know
the influencers on
Twitter
Build and Practice
• Work on your craft
a few hours
everyday