0% found this document useful (0 votes)

10 views58 pages

WK01 Lect-1

The document outlines the course structure for Engineering Vibration (MAE7015B) taught by Dr. Jian-Ping Li at the University of Bradford, including the importance of statistics in engineering, teaching organization, and assessment methods. It emphasizes the need for statistical knowledge in various engineering applications such as data analysis and quality control. The course includes lectures, tutorials, lab sessions, and assessments, with a focus on understanding variability and its impact on engineering design.

Uploaded by

Youssef Mahmoud

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

10 views58 pages

WK01 Lect-1

Uploaded by

Youssef Mahmoud

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 58

ENM5005B

STATISTICS and
Machine Learning
Dr. Jian-Ping Li
C0.15b/ Chesham
Tel. 4539 Email: [email protected]

1
University of Bradford Home Page

Engineering Vibration (MAE7015B)

After 10 minutes, students are not

allowed to enter classrooms.
Contents

1. Introduction of My Research
2. The reason to learn this module
3. Teaching Organisation
• Recommended reading
• Lectures
• Tutorials
• Lab organisation
4. Assessment
5. Support
6. Introduction of Contents
3
University of Bradford Home Page

Dr Jian-Ping Li

PhD Mechanical Engineering – University of

Manchester
http://www.scholarpark.co.uk/bradford/s/jpli
Research Interests:
❑ Engineering Modelling and Optimization
❑ Optimization theory and algorithms: Gradient-
based and evolutionary computation
opt_1.JPG (18768 bytes)

manchester.gif (2112 bytes) manchester.gif (2112 bytes)

Cambridge University
Engineering Vibration (MAE7015B)

Species Conserving Genetic Algorithm

Shubert Functions
n 5

Math Functions f (x) =

 j cos[( j + 1)x + j]
i =1 j =1
i Engineering
shubrt2r.gif (3528 bytes)

0.8

0.6

0.4

0.2
10
0
0 0.2 0.4 0.6 0.8 1 8
(#1)
6
-186~-184
4 -123~-121
-79~-78
2 -54~-50
-49~-44
0 -39~-30
-24~-19
-2 -17~-10
-9~-5
-4 >0.4

-6

-8

-10 3L
-10 -8 -6 -4 -2 0 2 4 6 8 10
1 3L
2
2D Shubert Function →18 global
6
solutions and 760 local solutions 3L 5 4L

10
Trap functions 3
4
200 9
160
z 4L
Fitness

200

160 120
Fitness

120
80
80
7
x y
40
40 8L
0 0
0 5 10 15 20
c 0 10 20 30
c 8L
8 L=0.635m(25in)

http://www.scholarpark.co.uk/mos
1. Why to learn Statistics

1. Big data science

2. Artificial Intelligent: Advertisement, voucher
3. In Engineering
1. Data analysis
2. Verification of design
3. Quality control

Attending lectures are very important for assessments

6
2. Teaching Organisation
Recommended Textbooks
Earlier editions are equally good

• Montgomery, D.C., Runger, G.C., Hubele,

N.F., Engineering Statistics, 5th Edition,
2012, J Wiley & Sons. [including student
solutions manual!]
• Vining, G., Kowalski, S.M., Statistical
Methods for Engineers, 2nd Edition, 2006,
Thomson Brooks/Cole.
• Morrisons, S.J., Introduction to Engineering
Statistics, 2012, Wiley [available as eBook]

Full reading list available in Canvas

7
2. Teaching Organisation
Teaching and Learning Methods
• Lectures
• Weeks 1-11, Thursday 9-10 am
• Notes will be made available in Canvas, recommended
that you have these available at the lecture!
• Problem-based teaching and assessment
• Teaching, tutorials and exam questions set in an
engineering context

Lecture notes will not be sufficient!

You will need to supplement with textbook based
study (reading + case studies)
8
2. Teaching Organisation
Teaching and Learning Methods
• Tutorials
• Weeks 2-11, Thursday 10-11 am
• Objective: practice statistical problem solving in class
(40-50% of the lecture time)
• Tutorial sheets available in Canvas before the lecture -
starting with Week 2
• Solutions for the “standard” tutorial problems will be
posted on Canvas 3-4 weeks after the tutorial.

• You will need calculators!!!

• Additional staff aviabliable to support tutorials
9
2. Teaching Organisation
Teaching and Learning Methods
Lab Sessions
• Objective: develop practical skills for using specialist software
(Minitab) to solve statistical engineering problems
• Thursday 11-12 Horton D1.04
• Friday 14-15, Horton D1.03
• Friday 17-18, Richmond L24
• Scheduled in smaller groups
• Check your timetable!!!

10
2. Teaching Organisation

Teaching and Learning Methods

Lectures
Time Room Demonstrator
9-10/Thursday B4.02 JP Li

Tutorial Sessions
Time/Date Room Program Demonstrator
10-11 am/ Thursday B4.02 All Zeixion Jin
/JP Li
11
3. Assessment

• Exam
• 70% weight Statistical formulae and tables provided

• Computer based test

• 30% weight
• Based on Minitab, open book – engineering problem

• Lab Time (10-11am,

10/4/2025,
Richmond F41/42)

12
General Support

• Previous exam papers available

• All course material and information will be
posted in Canvas
• You are welcome to email me
([email protected]) with any questions or
queries (but check Canvas first!)

13
Introduction to
Engineering Statistics

14
Introduction to Engineering Statistics

• Why is Statistics important to engineers?

https://www.youtube.com/watch?
v=886ahjgAxI8

AI = Statistics + big data

15
Introduction to Engineering Statistics

• Why is Statistics important to engineers?

In the words of one of a previous ES medical engineering students, while doing an
industrial placement:
“First of all, I would like to thank you for the Statistics module which I have
attended last academic year as part of my degree program.
At the beginning I couldn't really understand why should engineers bother
learning statistics as there will always be a statistician to help out. However,
I've realised all the significance of the knowledge you gave me in first few
weeks after commencing my placement. It
is used literally
everywhere - in testing, sample selection, data analysis
and even crucial decision making!
Knowing at least something about statistics means that you can interpret
statisticians’ data and sometimes even argue it.”
16
Why do we need Statistics?
Variability is the most fundamental law of nature!

17
Why do we need Statistics?
• Variability is the most fundamental law of nature;
• Variability affects performance of all systems;
• We need to take variability into account in all
aspects of engineering:

18
Why do we need Statistics?

• Variability is the most fundamental law of nature;

• Variability affects performance of all systems;
• We need to take variability into account in all
aspects of engineering:
• Engineering Analysis – variability in the input and in the
system itself will determine variability of the output;
• Engineering Science – we need laws and models that help us
to explain the actual behaviour of systems – including the
observed variability in the response (y);
• The success criteria for an engineering design / product is
minimal variation in functional performance.

19
How do you define “quality” of a
product?

• Quality is ultimately
judged on the basis of the
observed behaviour;
people always expect
consistent performance of
function around the ideal
/ target.
• Functions/Performance,
reliability, maintainability,
cost …
20
How do you define “quality” of a
product?

Cost

Performance / reliability

21
The Engineered System Model
Where do these “Noise Variables”
come from?

Variables / factors over which

the engineer has no control

Noise Variables

Engineered
Input System / Output
Product /
Process

Controlled
Variables
Engineering design parameters 22
The Engineered System Model
Change
Customer / over time
Systems
operation/ (e.g. wear)
interactions
use
Manufacturing
/ part to part
Environment
Noise Variables

Engineered
Input System / Output
Product /
Process

Controlled
Variables
Engineering design parameters 23
Statistical Thinking

• Statistical techniques are useful for describing

and understanding variability.
• By variability, we mean successive observations
of a system or phenomenon do not produce
exactly the same result.
• Statistics gives us a framework for describing
variability, and making engineering decisions
based on uncertainty associated with variability.

24
Engineering Statistics

Engineering Statistics supports the engineer in

• Planning and collecting engineering data;
• Presentation of data and formulation of
hypotheses;
• Analysis of data, and development of engineering
models / laws of the engineered systems;
• Use of data to make engineering decisions for
product and process design, and problem solving.

26
Engineering Statistics

Objectives of the module:

• To develop an understanding of variability
underpinning any engineering experiment;
• To develop knowledge and skills in
• Planning and collecting engineering data;
• Presentation of data and formulation of hypotheses;
• Analysis of data, and development of engineering models /
laws of the engineered systems;
• Use of data to make engineering decisions for product and
process design, and problem solving.

27
Topics Covered in the ES Module

Modelling Random Behaviour

Exploratory Data Analysis

28
Topics Covered in the ES Module

Weeks 1-4: ES Foundation Topics

• Exploratory Data Analysis
• Numerical and graphical representation of variability in
data
• Modelling random behaviour
• Use probability distributions (Normal / Gaussian
distribution in particular) to derive models that enable
us to predict future behaviour

29
ES Foundation Topics – Example
Problems
Example (1): Concrete strength
• A manufacturer of concrete structures is carrying out
concrete compressive strength tests on samples
collected from ongoing production at regular times.
The nominal specification is 60MPa.
• Sample of strength test results:
Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Strength 64.5 66 66.8 67 67.7 67.7 67.8 68.2 68.7 69 69 69.9 70.5 71.5 71.9
[Mpa]

Question: Can this supplier meet the

specification? 30
ES Foundation Topics – Example
Problems
Example (1): Concrete strength
Graphical
summary Numerical
(histogram) of characterisation
variation in of variation in
concrete concrete
strength strength

Normal
distribution
model for
concrete
strength
Predictions for the
mean concrete
strength
31
ES Foundation Topics – Example
Problems

Example 2: Customer Evaluation of

Opening Effort
• A customer evaluation clinic has been
set up to evaluate opening efforts for a
Fuel Filler Door.
• 50 customers were asked to rank the
performance of five concepts (on a
scale from 1 to 10) .

32
ES Foundation Topics – Example
Problems
Example 2: Customer Evaluation of Opening Effort
Response (1-10)
Competitor Competitor Prototype Prototype Prototype
C1 (22N) C2 (3N) P1 (4.5N) P2 (10.5N) P3 (18N)
1 4 8 7 9 9
2 4 5 6 9 6
3 5 4 8 9 9
4 7 4 9 10 9
5 7 4 8 8 9
6 8 5 9 9 8
7 3 5 8 10 6
8 5 3 6 9 8
9 7 4 6 9 6
10 4 5 9 8 8
and so on… Which concept is best?
(from a customer viewpoint) 33
Topics Covered in the ES Module

RSM
Design of
Experiments

Linear Regression
(fitting empirical models)

Statistical Process Control

Estimation & Testing

(Production Control, R&D Testing)

Modelling Random Behaviour

Exploratory Data Analysis and Machine Learning

34
What is Machine Learning?

“Learning is any process by which a system

improves performance from experience.” -
Herbert Simon.

Definition by Tom Mitchell (1998):

Machine Learning is the study of algorithms that

❑ improve their performance P

❑ at some task T
❑ with experience E.

35
36
Autonomous Car

37
Types of Learning

❑Supervised (inductive) learning

• Given: training data + desired outputs
(labels)
❑Unsupervised learning
• Given: training data (without desired
outputs) •
❑Reinforcement learning
• Rewards from sequence of actions

38
Supervised Learning: Regression

❑Given (x1, y1), (x2, y2), ..., (xn, yn)

❑Learn a function f(x) to predict y given x
– y is real-valued == regression

39
Supervised Learning: Classification

❑Given (x1, y1), (x2, y2), ..., (xn, yn)

❑Learn a function f(x) to predict y given x
– y is categorial == regression

40
Unsupervised Learning

Unsupervised learning is a machine learning technique

that uses algorithms to analyze and learn patterns from
unlabeled data without human intervention

41
Reinforcement Learning

Reinforcement learning (RL) is a machine learning (ML)

technique that teaches software to make decisions by
learning from its own experiences.

42
Reinforcement Learning

Reinforcement learning (RL) is a machine learning (ML)

technique that teaches software to make decisions by
learning from its own experiences.

43
Machine Learning vs Artificial Intelligence

44
EXPLORATORY DATA ANALYSIS

Part 1: Introduction

45
Exploratory Data Analysis
Outline Learning Objectives

• Types of data • Understand different types of engineering

data;
• Describing data with • Compute and interpret the sample mean,
numbers – variance, standard deviation, median, range
descriptive statistics and quartiles;
• Graphical methods • Understand the concepts of sample mean /
variance Vs. population mean / variance;
for describing data –
• Construct and interpret visual data displays,
construction and including histograms, stem-and-leaf
interpretation. diagrams, box-plots, run charts;
• Use data displays to identify trends, patterns
and features which could point to sources of
variability, and to compare two or more
samples of data.
46
Population Vs Sample
Following repeated customer complaints on
the charge capacity of supplied batteries, we
decided to test the batteries in stock. http://tbn0.google.com/images?q=tbn:NQbtKn6w7jxJoM:http://www.tayna.co.uk/getimage.php%3Fimage%3Dtayna/5368570146tayna.jpg