Learning Objective: Increase professional effectiveness, data management, and analytical skills
With evolving technology, many people are overloaded and overwhelmed with information and data. Businesses now have access to large amounts of feedback from internal and external sources. How do we make sense of the all of the information? Is the data reliable? How can we manage and utilize the data in order to impact business goals, visions, mission? This seminar with help you turn your information overload into powerful and reliable data that you can use to meet organizational goals.
At the end of this seminar, participants will be able to:
a. Assess and categorize data and information.
b. Identify tools and techniques to organize and interpret data.
c. Explore productivity tools and techniques.
d. Examine common data management challenges and solutions.
4. Activity
Rules
• The ball cannot be passed to the person next to you.
• The ball cannot be returned to the person who passed
it to you.
• No two persons can touch the ball at the same time.
• The ball must be released between passes.
• The team must repeat the cycle 3 times.
9. Dealing with Chaos
• Mindset
• Ability
• Agility
• Perspective
• Remove Barriers
• Paradigm Shift
• Sense of Purpose
• Clarity
10. How does Chaos look?
• Linear Regression
• With Outliers
• Without Outliers
11. Chaos in Character Recognition
● The word “Judge” was recognized as
“Yudge”
● “South Carolina 1900 - 1914”was
recognized as “South Carolinaogoo -
1914"
● “educational opportunities” was
recognized as “educational tp
portunities”
● “Mays” was recognized as as
“NI ays”
● “people’s lives” was recognized as
“Ito tple's lives”
● “South Carolina” was recognized as
“South :arolina”
● “South Carolina” was recognized as
“South (;arolina”
12. Structured Data vs. Unstructured Data
We generally think of Structured Data as having some known model or format.
This means that it can easily be broken down into smaller segments to look
for patterns. Unstructured Data does not generally fit this format.
Chaotic Data
• Text / tweets/ messages
• Audio / Video / Images
• Websites
13. • Reduce Complexity
– Simplify the data
• counts, grouping, parsing, understanding relationships,
distinguishing importance
– Rephrase your question
• What level of understanding do you need?
• What algorithms can you run on this data?
• Where does this get you?
• What work has been done on similar problems?
– Restructure the format
• graphs, network diagrams, decision trees
Dealing with Chaos – Build a Model
14. Dealing with Unstructured Data
Internet Websites
.Google’s PageRank algorithm found a way of
ranking the relevance of these sites based on links to
and from that site and ignored much of the site’s
other contents.
15. • Computer recognition for Understanding /
Comparison
– Latent Dirichlet Allocation Algorithm
• helps split a document into sections
• Spam Detection
– Naive Bayes Algorithm
• filters spam from legit emails
Dealing with Unstructured Data
16. • The Apriori Algorithm
– Used to discover items that occur together frequently in a set of transactions,
and can also create behavior rules for when these items will occur together.
• K-Means Clustering
– Based on a distance concept, and can be useful in grouping closer items
together forming K distinct groups.
• Markov Chains
– A model where transitions between states are only based on the present
state.
• Graph Theory
– Edges represent a relationship between two entities. This can be used to
model abstract relationships and use a wide body of literature on detecting
patterns in these graphs.
Machine Learning / Mining Tools
18. Taming the Data
• Vocalize
• Big Three Questions
• Validate
• Approach
• eValuate
• Tools and Techniques
“Pick battles big enough to matter but small enough to win.”
Jonathan Kozol
19. Big Three Questions
Question #1: What is not happening or working?
Describe the problem, issue or opportunity factually and objectively.
Question #2: Why do you say that?
State what objective evidence has been found to prove the problem,
issue or opportunity exists.
Question #3: So what?
Estimate the potential financial impact of the problem, issue or
opportunity on the organization and/or customer.
21. Validate Approach
ProjectDurationandComplexity
Six Sigma (DFSS / DMADV)
Large scale; New process design /
re-engineering
Six Sigma (DMAIC)
Larger scale, more complex projects;
Root cause unknown
LEAN
Reduce waste, cycle time and complexity;
Increase consistency; Standardize
Just-Do-It
Logic/intuition
23. Six Sigma
Six Sigma - a disciplined, data-driven approach and
methodology for eliminating defects
24. Six Sigma
Six Sigma is measure of variability equal to 3.4 defects
per million opportunities.
Activity 99% Defect Free (3.8 Sigma) 99.9997% Defect Free (6 Sigma)
Mail Delivery
20,000 lost articles of mail per
hour.
Seven lost articles of mail per hour.
Drinking Water
Unsafe drinking water for 15
minutes each day.
Unsafe drinking water for two
minutes per year.
Hospital Surgery
5,000 incorrect procedures each
week.
Two incorrect procedures each
week.
Air Travel
Two abnormal landings at most
airports each day.
One abnormal landing at most
airports every five years.
Drug
Prescription
200,000 wrong drug prescriptions
each year.
68 wrong drug prescriptions each
year.
26. Six Sigma – Key Methodologies
Both result in more predictable, mature and
defect free performance
27. eValuate - DMAIC Tools
Affinity
Diagram
Brainstorming
Drill-Down
Tree
Focus Groups
In/Out of the
Frame
SIPOC
AnalyzeMeasureDefine Improve Control
Benchmarking
Gage R&R
KANO
Analysis
Pareto Chart
RACI Chart
Spaghetti
Diagram
4-Blocker
ANOVA
Correlation and
Regression
Graphs and
Plots
Hypothesis
Testing
Ishikawa
Diagram
Design of
Experiments
Mistake
Proofing
Failure Mode
and Effects
Analysis
Pugh Matrix
SCAMPER
TRIZ
Implementation
and Control Plan
C / U Chart
I-MR Chart
Metrics
NP / P Chart
Xbar R / S Chart
Process variation is reduced by determining and controlling
the vital inputs that cause variation in the output(s).
.
28. Affinity Diagram
• Sort large amounts of information
• Create meaningful categories
• Determine primary groups
• Encourage nontraditional connections
29. Affinity Diagram
Step 1: Phrase the issue.
Step 2: Brainstorm silently and
record ideas on large post-it notes.
- Generate at least 20 ideas.
- Include a noun and a verb and limit
idea to less than 7 words.
- Use large print (ideas should be visible
from 4 to 6 feet away).
Step 3: Display the completed
cards/post-it notes on the wall or
conference table.
Low Cost
Airfares
Leather Seats
On Time
Departures
No Luggage
Lost
Safe Flight
Ample Leg
Room
Direct Flight
Free Head
Phones
Clean
Restrooms
What are the key outputs for
airline passengers?
Warm Cookies
30. Affinity Diagram
Online tools: stormboard and powernoodle
Low Cost AirfaresLeather Seats Warm Cookies
Safe Flight Free Head PhonesOn Time Departure
Ample Leg Room Direct Flights
What are the key outputs for airline passengers?
Comfort / Safety Competitiveness Service
No Luggage Lost
Clean Restrooms
Step 4: Sort into related groupings and define the group headers
(Sort in silence; let groupings emerge naturally)
Step 5: Finished product - Affinity diagram
31. Pareto Chart
• Pareto principle
• Prioritize elements
• Large number of elements
• High cost to evaluate
• Potential root causes
• Visual justification
A minority of input produces
the majority of results.” –
Vilfredo Pareto
32. Pareto Chart
Count 343 30 26 8 6 5 5 4
Percent 2.333.1 23.1 20.0 6.2 4.6 3.8 3.8 3.1
Cum % 100.033.1 56.2 76.2 82.3 86.9 90.8 94.6 97.7
Error Type
O
ther
m
issed
pop
fly
in
infield
w
ild
pitch
m
issed
coaching
signal
failure
to
back
up
the
play
m
issed
grounder
throw
to
w
rong
cutoff player
m
issed
fly
ball in
outfield
taken
third
strike
140
120
100
80
60
40
20
0
100
80
60
40
20
0
Count
Percent
Pareto Chart of Error Type
33. Cause and Effect / Fishbone Diagram
Ishikawa Diagram
Focuses the team on causes, not symptoms
or the history of a problem or failure.
36. DMAIC Tools
Affinity
Diagram
Brainstorming
Drill-Down
Tree
Focus Groups
In/Out of the
Frame
SIPOC
AnalyzeMeasureDefine Improve Control
Benchmarking
Gage R&R
KANO
Analysis
Pareto Chart
RACI Chart
Spaghetti
Diagram
4-Blocker
ANOVA
Correlation and
Regression
Graphs and
Plots
Hypothesis
Testing
Ishikawa
Diagram
Design of
Experiments
Mistake
Proofing
Failure Mode
and Effects
Analysis
Pugh Matrix
SCAMPER
TRIZ
Implementation
and Control Plan
C / U Chart
I-MR Chart
Metrics
NP / P Chart
Xbar R / S Chart
37. Pugh Matrix
Task: To reduce the
time required to enter
cost data into the
accounting system
Customer or business
requirements
Importance ranking can come from a previously completed
Quality Function Deployment or Cause & Effect Matrix with Prioritization
38. Pugh Matrix
(+) Positive impact
compared to the baseline.
(-) Negative impact
compared to the baseline.
(s) Same impact compared
to the baseline.
40. TRIZ - Theory of Inventive Problem Solving
Established methodology for stimulating and generating
innovative ideas and solutions for problem solving
41. TRIZ - Theory of Inventive Problem Solving
Traditional Problem Solving vs. TRIZ
Problem
Analysis
Abstract
Problem
Abstract
Solution
Specific
Problem
Specific
Solution
Concept
4
Concept
3
Concept
2
Concept
1
46. Challenge
In order to increase profitability and reach the 2020 goal of $50M in net
revenue, company X must decrease overheads. Preliminary results show
that reproduction costs have been on the rise over the years and have
increased the overhead burden rate by 5% per year from $8.52 per man-
hour to $10.87 over the last 5 years. In order to make a copy
professionals must type in a project and phase into the copy machine or
create a requisition with the in-house Reproduction Department, spending
an average an additional 30mins per day on this activity. Even with the
inconvenience, professionals consume the equivalent of ½ ream of paper
per day, 60% of which are color copies or prints which cost $0.50 per
sheet vs $0.05 per sheet for black and white prints.
Your challenge is to reduce reproduction costs.
47. Reliable Business Results
• Deal with the Chaos
– Begin with a Clear Mind
• Tame the Data
– Vocalize, Validate, eValuate
• Own the Results
– Improve, Control, Manage