Admin
مدير المنتدى
عدد المساهمات : 18996
التقييم : 35494
تاريخ التسجيل : 01/07/2009
الدولة : مصر
العمل : مدير منتدى هندسة الإنتاج والتصميم الميكانيكى
 |  موضوع: كتاب MATLAB Machine Learning Recipes - A Problem-Solution Approach الجمعة 17 مايو 2024, 12:19 am | |
| 
أخواني في الله
أحضرت لكم كتاب
MATLAB Machine Learning Recipes - A Problem-Solution Approach
Third Edition
Michael Paluszek , Stephanie Thomas
و المحتوى كما يلي :
Contents
About the Authors XVII
About the Technical Reviewer XIX
Introduction XXI
1 An Overview of Machine Learning 1
1.1 Introduction . 1
1.2 Elements of Machine Learning . 2
1.2.1 Data 2
1.2.2 Models 2
1.2.3 Training 3
1.3 The Learning Machine . 4
1.4 Taxonomy of Machine Learning 6
1.5 Control . 8
1.5.1 Kalman Filters 9
1.5.2 Adaptive Control . 9
1.6 Autonomous Learning Methods 10
1.6.1 Regression . 10
1.6.2 Decision Trees 13
1.6.3 Neural Networks . 14
1.6.4 Support Vector Machines (SVMs) 16
1.7 Artificial Intelligence 17
1.7.1 What Is Artificial Intelligence? 17
1.7.2 Intelligent Cars 17
1.7.3 Expert Systems 18
1.8 Summary 19
2 Data for Machine Learning in MATLAB 21
2.1 Introduction to MATLAB Data Types . 21
2.1.1 Matrices 21
2.1.2 Cell Arrays 22
2.1.3 Data Structures 23
2.1.4 Numerics . 25
2.1.5 Images . 25
VCONTENTS
2.1.6 Datastore . 26
2.1.7 Tall Arrays 28
2.1.8 Sparse Matrices 29
2.1.9 Tables and Categoricals 30
2.1.10 Large MAT-Files . 32
2.2 Initializing a Data Structure . 33
2.2.1 Problem 33
2.2.2 Solution 33
2.2.3 How It Works . 33
2.3 mapreduce on an Image Datastore . 36
2.3.1 Problem 36
2.3.2 Solution 36
2.3.3 How It Works . 36
2.4 Processing Table Data 38
2.4.1 Problem 38
2.4.2 Solution 38
2.4.3 How It Works . 39
2.5 String Concatenation 42
2.5.1 Problem 42
2.5.2 Solution 42
2.5.3 How It Works . 42
2.6 Arrays of Strings 42
2.6.1 Problem 42
2.6.2 Solution 42
2.6.3 How It Works . 42
2.7 Substrings 43
2.7.1 Problem 43
2.7.2 Solution 43
2.7.3 How It Works . 43
2.8 Reading an Excel Spreadsheet into a Table . 44
2.8.1 Problem 44
2.8.2 Solution 44
2.8.3 How It Works . 44
2.9 Accessing ChatGPT . 46
2.9.1 Problem 46
2.9.2 Solution 46
2.9.3 How It Works . 46
2.10 Summary 48
VICONTENTS
3 MATLAB Graphics 49
3.1 2D Line Plots 49
3.1.1 Problem 49
3.1.2 Solution 49
3.1.3 How It Works . 50
3.2 General 2D Graphics 52
3.2.1 Problem 52
3.2.2 Solution 52
3.2.3 How It Works . 53
3.3 Custom Two-Dimensional Diagrams 54
3.3.1 Problem 54
3.3.2 Solution 54
3.3.3 How It Works . 55
3.4 Three-Dimensional Box . 56
3.4.1 Problem 56
3.4.2 Solution 56
3.4.3 How It Works . 56
3.5 Draw a 3D Object with a Texture 59
3.5.1 Problem 59
3.5.2 Solution 59
3.5.3 How It Works . 60
3.6 General 3D Graphics 61
3.6.1 Problem 61
3.6.2 Solution 61
3.6.3 How It Works . 62
3.7 Building a GUI . 63
3.7.1 Problem 63
3.7.2 Solution 63
3.7.3 How It Works . 63
3.8 Animating a Bar Chart . 68
3.8.1 Problem 68
3.8.2 Solution 69
3.8.3 How It Works . 69
3.9 Drawing a Robot 73
3.9.1 Problem 73
3.9.2 Solution 73
3.9.3 How It Works . 73
3.10 Importing a Model . 76
3.10.1 Problem 76
3.10.2 Solution 76
3.10.3 How It Works . 76
3.11 Summary 83
VIICONTENTS
4 Kalman Filters 85
4.1 Gaussian Distribution 86
4.2 A State Estimator Using a Linear Kalman Filter 87
4.2.1 Problem 87
4.2.2 Solution 88
4.2.3 How It Works . 89
4.3 Using the Extended Kalman Filter for State Estimation 106
4.3.1 Problem 106
4.3.2 Solution 107
4.3.3 How It Works . 108
4.4 Using the UKF for State Estimation 111
4.4.1 Problem 111
4.4.2 Solution 112
4.4.3 How It Works . 113
4.5 Using the UKF for Parameter Estimation 117
4.5.1 Problem 117
4.5.2 Solution 118
4.5.3 How It Works . 118
4.6 Range to a Car 122
4.6.1 Problem 122
4.6.2 Solution 122
4.6.3 How It Works . 122
4.7 Summary 125
5 Adaptive Control 127
5.1 Self-Tuning: Tuning an Oscillator . 128
5.1.1 Problem 129
5.1.2 Solution 130
5.1.3 How It Works . 130
5.2 Implement MRAC 136
5.2.1 Problem 136
5.2.2 Solution 136
5.2.3 How It Works . 136
5.3 Generating a Square Wave Input 140
5.3.1 Problem 140
5.3.2 Solution 140
5.3.3 How It Works . 140
5.4 Demonstrate MRAC for a Rotor 142
5.4.1 Problem 142
5.4.2 Solution 142
5.4.3 How It Works . 142
VIIICONTENTS
5.5 Ship Steering: Implement Gain Scheduling for Steering Control of a Ship 145
5.5.1 Problem 145
5.5.2 Solution 145
5.5.3 How It Works . 145
5.6 Spacecraft Pointing . 148
5.6.1 Problem 148
5.6.2 Solution 148
5.6.3 How It Works . 148
5.7 Direct Adaptive Control . 153
5.7.1 Problem 153
5.7.2 Solution 153
5.7.3 How It Works . 153
5.8 Summary 155
6 Fuzzy Logic 157
6.1 Building Fuzzy Logic Systems . 158
6.1.1 Problem 158
6.1.2 Solution 158
6.1.3 How It Works . 158
6.2 Implement Fuzzy Logic . 163
6.2.1 Problem 163
6.2.2 Solution 163
6.2.3 How It Works . 163
6.3 Window Wiper Fuzzy Controller 169
6.3.1 Problem 169
6.3.2 Solution 169
6.3.3 How It Works . 169
6.4 Simple Discrete HVAC Fuzzy Controller 174
6.4.1 Problem 174
6.4.2 Solution 174
6.4.3 How It Works . 174
6.5 Variable HVAC Fuzzy Controller 180
6.5.1 Problem 180
6.5.2 Solution 181
6.5.3 How It Works . 181
6.6 Summary 189
7 Neural Aircraft Control 191
7.1 Longitudinal Motion 191
7.1.1 Problem 193
7.1.2 Solution 193
7.1.3 How It Works . 193
IXCONTENTS
7.2 Numerically Finding Equilibrium 198
7.2.1 Problem 198
7.2.2 Solution 199
7.2.3 How It Works . 199
7.3 Numerical Simulation of the Aircraft 200
7.3.1 Problem 200
7.3.2 Solution 200
7.3.3 How It Works . 201
7.4 Activation Function . 202
7.4.1 Problem 202
7.4.2 Solution 203
7.4.3 How It Works . 203
7.5 Neural Net for Learning Control 204
7.5.1 Problem 204
7.5.2 Solution 204
7.5.3 How It Works . 204
7.6 Enumeration of All Sets of Inputs . 208
7.6.1 Problem 208
7.6.2 Solution 208
7.6.3 How It Works . 208
7.7 Write a Sigma-Pi Neural Net Function . 210
7.7.1 Problem 210
7.7.2 Solution 210
7.7.3 How It Works . 210
7.8 Implement PID Control . 213
7.8.1 Problem 213
7.8.2 Solution 213
7.8.3 How It Works . 213
7.9 PID Control of Pitch 218
7.9.1 Problem 218
7.9.2 Solution 218
7.9.3 How It Works . 218
7.10 Neural Net for Pitch Dynamics . 223
7.10.1 Problem 223
7.10.2 Solution 223
7.10.3 How It Works . 224
7.11 Nonlinear Simulation 226
7.11.1 Problem 226
7.11.2 Solution 226
7.11.3 How It Works . 226
7.12 Summary 228
XCONTENTS
8 Introduction to Neural Nets 229
8.1 Daylight Detector 229
8.1.1 Problem 229
8.1.2 Solution 229
8.1.3 How It Works . 230
8.2 Modeling a Pendulum 231
8.2.1 Problem 231
8.2.2 Solution 232
8.2.3 How It Works . 232
8.3 Single Neuron Angle Estimator . 235
8.3.1 Problem 235
8.3.2 Solution 236
8.3.3 How It Works . 236
8.4 Designing a Neural Net for the Pendulum . 240
8.4.1 Problem 240
8.4.2 Solution 240
8.4.3 How It Works . 240
8.5 XOR Example 244
8.6 Training . 253
8.7 Summary 254
9 Classification of Numbers Using Neural Networks 257
9.1 Generate Test Images with Defects . 257
9.1.1 Problem 257
9.1.2 Solution 258
9.1.3 How It Works . 258
9.2 Create the Neural Net Functions 262
9.2.1 Problem 262
9.2.2 Solution 263
9.2.3 How It Works . 263
9.3 Train a Network with One Output Node 267
9.3.1 Problem 267
9.3.2 Solution 267
9.3.3 How It Works . 269
9.4 Testing the Neural Network . 272
9.4.1 Problem 272
9.4.2 Solution 272
9.4.3 How It Works . 273
9.5 Train a Network with Many Outputs 273
9.5.1 Problem 273
9.5.2 Solution 273
9.5.3 How It Works . 274
9.6 Summary 277
XICONTENTS
10 Data Classification with Decision Trees 279
10.1 Generate Test Data . 280
10.1.1 Problem 280
10.1.2 Solution 280
10.1.3 How It Works . 280
10.2 Drawing Trees 283
10.2.1 Problem 283
10.2.2 Solution 283
10.2.3 How It Works . 283
10.3 Implementation . 287
10.3.1 Problem 287
10.3.2 Solution 287
10.3.3 How It Works . 288
10.4 Creating a Tree . 291
10.4.1 Problem 291
10.4.2 Solution 291
10.4.3 How It Works . 291
10.5 Handmade Tree . 295
10.5.1 Problem 295
10.5.2 Solution 295
10.5.3 How It Works . 295
10.6 Training and Testing 298
10.6.1 Problem 298
10.6.2 Solution 298
10.6.3 How It Works . 298
10.7 Summary 301
11 Pattern Recognition with Deep Learning 303
11.1 Obtain Data Online for Training a Neural Net . 305
11.1.1 Problem 305
11.1.2 Solution 305
11.1.3 How It Works . 305
11.2 Generating Training Images of Cats 305
11.2.1 Problem 305
11.2.2 Solution 305
11.2.3 How It Works . 306
11.3 Matrix Convolution . 308
11.3.1 Problem 308
11.3.2 Solution 309
11.3.3 How It Works . 309
11.4 Convolution Layer . 311
11.4.1 Problem 311
XIICONTENTS
11.4.2 Solution 311
11.4.3 How It Works . 311
11.5 Pooling to Outputs of a Layer 312
11.5.1 Problem 312
11.5.2 Solution 312
11.5.3 How It Works . 313
11.6 Fully Connected Layer . 314
11.6.1 Problem 314
11.6.2 Solution 314
11.6.3 How It Works . 314
11.7 Determining the Probability . 316
11.7.1 Problem 316
11.7.2 Solution 316
11.7.3 How It Works . 317
11.8 Test the Neural Network 318
11.8.1 Problem 318
11.8.2 Solution 318
11.8.3 How It Works . 318
11.9 Recognizing an Image . 320
11.9.1 Problem 320
11.9.2 Solution 320
11.9.3 How It Works . 320
11.10 Using AlexNet . 322
11.10.1 Problem 322
11.10.2 Solution 322
11.10.3 How It Works . 322
Summary 326
12 Multiple Hypothesis Testing 327
12.1 Overview 327
12.2 Theory 329
12.2.1 Introduction 329
12.2.2 Example 331
12.2.3 Algorithm . 331
12.2.4 Measurement Assignment and Tracks 333
12.2.5 Hypothesis Formation . 334
12.2.6 Track Pruning . 335
12.3 Billiard Ball Kalman Filter . 335
12.3.1 Problem 335
12.3.2 Solution 336
12.3.3 How It Works . 336
12.4 Billiard Ball MHT 342
XIIICONTENTS
12.4.1 Problem 342
12.4.2 Solution 342
12.4.3 How It Works . 342
12.5 One-Dimensional Motion 345
12.5.1 Problem 345
12.5.2 Solution 346
12.5.3 How It Works . 347
12.6 One-Dimensional MHT . 349
12.6.1 Problem 349
12.6.2 Solution 349
12.6.3 How It Works . 349
12.7 Summary 351
13 Autonomous Driving with MHT 355
13.1 Automobile Dynamics . 356
13.1.1 Problem 356
13.1.2 Solution 356
13.1.3 How It Works . 356
13.2 Automobile Radar 359
13.2.1 Problem 359
13.2.2 Solution 359
13.2.3 How It Works . 359
13.3 Passing Control . 362
13.3.1 Problem 362
13.3.2 Solution 362
13.3.3 How It Works . 362
13.4 Automobile Animation . 363
13.4.1 Problem 363
13.4.2 Solution 364
13.4.3 How It Works . 364
13.4.4 Solution 364
13.5 Automobile Simulation and the Kalman Filter . 367
13.5.1 Problem 367
13.5.2 Solution 368
13.5.3 How It Works . 368
13.6 Automobile Target Tracking 371
13.6.1 Problem 371
13.6.2 Solution 371
13.6.3 How It Works . 371
13.7 Summary 374
XIVCONTENTS
14 Spacecraft Attitude Determination 377
14.1 Star Catalog . 377
14.1.1 Problem 377
14.1.2 Solution 378
14.1.3 How It Works . 378
14.2 Camera Model 381
14.2.1 Problem 381
14.2.2 Solution 381
14.2.3 How It Works . 381
14.3 Celestial Sphere . 383
14.3.1 Problem 383
14.3.2 Solution 383
14.3.3 How It Works . 383
14.4 Attitude Simulation of Camera Views . 384
14.4.1 Problem 384
14.4.2 Solution 384
14.4.3 How It Works . 384
14.5 Yaw Angle Rotation . 387
14.5.1 Problem 387
14.5.2 Solution 387
14.5.3 How It Works . 387
14.6 Yaw Images . 388
14.6.1 Problem 388
14.6.2 Solution 388
14.6.3 How It Works . 388
14.7 Attitude Determination . 391
14.7.1 Problem 391
14.7.2 Solution 391
14.7.3 How It Works . 391
14.8 Summary 399
15 Case-Based Expert Systems 401
15.1 Building Expert Systems 404
15.1.1 Problem 404
15.1.2 Solution 404
15.1.3 How It Works . 404
15.2 Running an Expert System . 406
15.2.1 Problem 406
15.2.2 Solution 407
15.2.3 How It Works . 407
15.3 Summary 410
XVCONTENTS
A A Brief History 411
B Software for Machine Learning 419
Bibliography 431
Index 435
Index
A
Activation functions, 202–203, 236, 237
Adaptive control
designed and implemented, 127
learning and adaptive, 127
self-tuning, 128
taxonomy, 127, 128
Adaptive control systems, xiii, 2, 4, 7, 9–10
Adaptive controllers, 416
Adaptive/intelligent control, 415, 416
Advice Taker, 412
Aerodynamic coefficients, 194
Aerodynamic models, 193, 198
Air traffic control radar systems, 327
Aircraft dynamics, 415
AircraftSim, 219–221, 226
AircraftSimOpenLoop, 201–202
AlexNet, 322–325, 421
Artificial intelligence (AI)
advances, 414
Advice Taker, 412
Bayesian network, 414
blocks world, 412
ChatGPT, 414
chess, 413, 414
definition, 17
expert systems, 18, 413
GPS, 412
Hanoi Towers, 412
intelligent cars, 17–18
limitations, 413
Lisp, 412
LT, 411
mathematical formulation, 411
neural network, 411
Time-sharing, 412
timeline, 414
Watson, 414
Artificial neural networks, 413
Aspect ratio, 193, 194
AttitudeDetermination function, 391, 399
AttitudeSim function, 384, 386, 387, 399
Automatic control systems, 2
Automobile animation
automobile 3D model, 364
AutomobileDemo, 366
cars, 363
DrawComponents, 365, 366
drawnow, 366
graphics window, 366
LoadOBJFile, 364
Macintosh application, 364
mesh, 365
.obj files, 364
OBJ model, 364
passing sequence, 366, 367
patch. patch, 365
updation, 365
Automobile controllers, 374
Automobile dynamics
collisions, 356
contact friction, 358
dynamic pressure, 357
dynamical equations, 357
inertial velocity, 359
planar model, 356, 357
RHSAutomobile function, 356
rolling friction, 358
RungeKutta function, 356
steering angle, 359
tire force, 358
transformation, 358
unit vector, 358
vehicle states, 356
velocity derivatives, 356
wheel force and torque, 358
yaw angle and yaw angular rate, 359
Automobile radar
AutoRadar function, 360
AutoRadarUKF function, 360–361
built-in radar demo, 361
delta velocity, 360
demos, 360
functions, 359
power, 359
radar model, 359
range, 359, 360
sensor, 361
Automobile simulation
automobile positions, 369
automobile track, 369
automobile trajectories, 370
car tracking, 367
demonstration, 368
differential equations, 368
Kalman Filter, 368
MHT distance, 368, 371
MHTDistanceUKF function, 368–369
radar measurements, 367
RHSAutomobileXY function, 368
true states and UKF estimated states, 370
UKFAutomobileDemo, 369
velocity, 368
Automobile target tracking
automobile demo, 374
demo car trajectories, 372
demo radar measurement, 373
IMM, 372
maneuvers, 372
MHT system, 371, 372
MHTAutomobileDemo, 371
primary car, 371
radar, 371
tracks, 373
UKF, 372
AutomobileDemo function, 375
AutomobileInitialize function, 375
AutomobileLaneChange function, 375
AutomobilePassing function, 375
Autonomous cars, xiii, 19
Autonomous control systems, 17
Autonomous driving, 417
car, 355
Kalman Filter, 355
measurements, 355
model, tracked automobiles, 355
passing control (see Passing control)
Autonomous learning, 6, 418
AutoRadar function, 375
AutoRadarUKF function, 375
B
Backpropagation, 266, 413
Batch process, 4
Bayes theorem, 93, 96, 417
Bayesian network, 18, 414
Big data, 419
Binary decision tree, 279, 283, 291
Binary trees, 279, 280, 283, 284, 286, 287
BuildFuzzySystem, 159–160, 169, 181
C
Camera model
patch, 382
pinhole camera, 381, 382
PinholeCamera function, 381
plot3, 382–383
star identification, 381
catColorMapper, 37
Categorical arrays, 30–31
CelestialSphere function, 383, 399
Cell arrays, 22–23
CFITSIO library, 25
changeFonts boolean, 261
436INDEX
Chapman-Kolmogorov equation, 97
ChatGPT, 8, 46–48, 414
Chess programs, 413, 414
C Language Integrated Production
System (CLIPS), 413
Classical ballet technique, 3
Classification methods, 420
Clustering algorithms, 3, 420
Combat aircraft, 418
Combinations, 208–210
Content generation systems, 2
Control systems, 8
adaptive, 2, 4, 7, 9–10
automatic, 2
autonomous, 17
definition, 2
feedback, 8
feedforward, 5
optimal, 7
Titan landing, 16
Convolutional neural network (CNN), 391, 392
AlexNet for image classification,
322–325
calculate probability, 316–317
code listing, 326
convolution-connected layer, 311–312
deep learning CNN, 303, 304
generate training images of cats, 305–308
implement fully connected layer, 314–316
matrix convolution, 308–311
online for training cat recognition neural
net, 305
pool layer outputs, 312–314
pooling layer, 303, 304
recognize image, 320–321
test neural network, 318–319
types of layers, 303
CreateDigitImage function, 258
Cumulative Probability Density Function
(CPDF), 86
Custom two-dimensional diagrams, 54–56
CVX, 425
D
Data, 2
mining, 414, 417
structure, 33–36
Data classification
code listing, 301
create tree, 291–295
draw binary decision tree, 283–287
generate test data, 280–283
handmade decision tree, 295–298
implement Gini impurity measure,
287–291
training and testing, 298–301
Daylight detector, 231
detector results, 230
light level measure, photocell, 230
problem, 229
solar flux, 230
solution, 229
Decision tree, 13, 409
binary trees, 279, 280, 283, 286, 287, 291
classification, 13–14
classify data using decision tree (see Data
classification)
MHT tree, 344, 345
Deep learning network GUI, 246
Deep neural network, 15
Defuzzification, 159, 162, 163, 166, 168
Delta acceleration (DA), 224, 226
DI, see Dynamic inversion (DI)
Digit0FontsTS, 268
DigitTrainingData, 260, 272
DigitTrainingTS, 269
Direct adaptive control, 153–155
Discrete HVAC fuzzy controller
arbitrary mode setting, 179
dynamical model, 174–176
heating system and air conditioning, 174
house model, 174, 175
HVACFuzzyPlot.m, 179–180
HVACSim.m, 176–177
initialize mode, 177–178
437INDEX
inputs and outputs, 180
non-fuzzy hysteresis controller
performance, 177
set of rules, 174, 175
simulation, 180, 181
temperature categories, 174, 175
variables, 174
Discriminative model, 16
DLXORNoisy.m, 251
Drag polar, 193
DrawComponents function, 375
dRHSL, 224, 226
Dynamic inversion (DI)
acceleration, 221
nonlinear, 192
vs. PID, 204, 221, 223
pitch, 218
E
Earth-Centered Inertial (ECI), 379, 380
EquilibriumState, 199–200
Excel spreadsheet, 44–45
Expert systems, xiii, 18, 41
AI, 413
case-based, 401, 402
CLIPS, 426
definition, 413
fact-gathering, 403
fully autonomous based reasoning, 403
new rule sets, 403
products, 426
rule-based, 401, 403, 404
Extended Kalman filter
AngleMeasurementPartial, 108
conventional, 111
EKFPredict and EKFUpdate functions,
109
EKFSim script implements, 110
nonlinear model, 106
solution, 107
state derivative function, 108
F
Face identification system, 3
Face recognition, 3, 5, 15, 25
Fast Fourier Transform (FFT), 9
Feedback control systems, 8
Feedforward control system, 5
feedforwardnet, 251
Flexible Image Transport System (FITS), 25
fminsearch, 199, 200
F-16 data, 198
F-16 model, 196–198
for loop, 260
Fuzzify, 164, 165
Fuzzy logic, 8
build systems
BuildFuzzySystem, 159–160
commercial/open source tools, 163
contains, 160
data structure, 158
defuzzification, 159
fuzzy rules, 162
fuzzy sets, 161
inference engine, 158
MATLAB function, 158
membership functions, 161–162
set and rule structure, 159
structure, 162
framework of set theory, 157
functions and scripts, code, 189, 190
helper functions, 189, 190
HVAC fuzzy controller
discrete, 174–181
variable, 180–189
implement, 163–168
TRLs, 157
window wiper fuzzy controller,
169–174
Fuzzy rules, 159, 162, 165, 168
Fuzzy sets, 158, 159, 161, 162
FuzzyInference, 164–168
FuzzyPlot, 169
FuzzyRand, 172
438INDEX
G
Gaussian distribution
CPDF and PDF, 86–87
Gaussian membership function, 161
Gaussian variable, 95, 98, 100
General Problem Solver (GPS), 412
Generate test images with defects
CreateDigitImage function, 258
digit neural network setup files, 262
DigitTrainingData, 260
digit training sets, 262
helper function, 261
for loop, 260
oneDigitMode, 261
Poisson or shot noise, 257
print, 258
SaveTS, 260
simple Poisson or shot noise, 257, 258
text, 258
Generative deep learning systems, 4
Generative machine learning (ML) models, 16,
418
Geosynchronous communications satellites, 416
GNU Linear Programming Kit (GLPK), xv, 422
H
Hessian matrix, 253
HiddenMarkov models, 413, 414
HVAC fuzzy controller
discrete, 174–181
variable, 180–189
HVACFuzzyController.m, 181–186
HVACFuzzyPlot.m, 179–180
HVACSim.m, 176–177
HVACSimplestFuzzyController, 174
I, J
imagesc, 270
Implement fuzzy logic
control flow, 164
defuzzification, 168
defuzzify, 163
function handles, 163
Fuzzify, 165
fuzzy rule logic, 165
FuzzyInference164–168
Inertia matrix, 197
Intelligent cars, 17–18
Interactive Multiple Models (IMM), 372, 422
K
Kalman filters, 9, 18, 207, 334, 355
acceleration, 86
application, 86
billiard ball Kalman Filter, 335–342
block diagram, 93
car tracking
dynamic model, 122
random walk, 125
state propagation, 124
family tree, 94
Gaussian distribution (see Gaussian
distribution)
implementation
constant acceleration, 102
MATLAB, 101
state space, 101
important, 85
mathematical framework, 85
MHT system, 342
noise matrix, 105–106
one-dimensional MHT, 349
one-dimensional motion, 346, 348
parameters, 85
prediction, 104
script, 105
simulation, 103
unscented, 94
update, 104–105
KFBilliardsDemo, 337–339, 341–342
Knowledge acquisition, 18
Knowledge-based systems, 413
kSigmoid, 211
439INDEX
L
Labeling, 3
Large MAT file, 32
Learning control, 415–416
Learning machine, 4–5
Levenberg-Marquardt training algorithm, 253,
254
LIBSVM, 423
Linear measurement, 9
Linear regression, 10, 11, 13
Lisp, 412
LoadHipparcos function, 378, 399
Logic Theorist (LT), 412
Longitudinal control, 191
Longitudinal motion
aerodynamic coefficients, 194
aerodynamic models, 198
aircraft dynamics symbols, 193, 194
differential equations, 193
dynamical equations, 194–195
F-16 data, 198
F-16 model, 196–197
inertia matrix, 197
learning control, 191–192
neural network, 192
PID controller, 192
pitch equation, 195
RHSAircraft, 195–196
LOQO, 424
M
Machine learning, 1, 411, 414, 417
autonomous driving, 417
autonomous learning software, 419
Bayes’ theorem, 417
data mining, 417
definition, 1
elements
data, 2
models, 2
training, 3–4
engineering, xiii
fusion reactors, xiii
GLPK, xv
machine vision, 417
MATLAB, xiii
MATLAB software (see MATLAB
software)
MATLAB strings (see MATLAB
stringdata type)
Non-MATLAB products, 422–423
optimization tools
CVX, 424
GLPK, 424
LOQO, 424
SeDuMi, 424
SNOPT, 424
uses, 423
YALMIP, 424
packages, xiii
PlotSet function, xiv–xv
SVMs, 417
taxonomy, 6–8
traditional, 6
two-dimensional arrays, 68
mapreduce, 36–38
Mathworks, 322
Statistics and Machine Learning Toolbox,
420
MATLAB dat string
arrays of strings, 42–43
categorical arrays, 30–31
large MAT file, 32
string concatenation, 42
substrings, 43
MATLAB function, 158
MATLAB graphics
animating a bar chart, 68–73
building GUI, 62–68
custom two-dimensional diagrams, 54–56
drawing a robot, 73–76
importing a model, 76–83
three-dimensional box, 56–59
3D graphics, 61–62
440INDEX
3D object with texture, 59–61
2D graphics, 52–54
2D line plots, 49–52
MATLAB mex files
cbkFunction, 428
Clang, 428
CLIPS, 426–429
facts, 428, 429
libCLIPS.dylib, 428
MEXTest.c., 427
MEXTest.h, 427–428
MEXTest.m, 428
Rules.CLP, 428
XCode, 428
MATLAB software
MathWorks products, 420
deep learning toolbox, 421
global optimization toolbox, 421
optimization toolbox, 421
statistics and machine learning toolbox,
420
text analytics toolbox, 421
Princeton Satellite Systems products
Core Control toolbox, 421
target tracking, 422
MATLAB stringdata type
cell arrays, 22–23
data structures, 22–25
datastores, 26–28
images, 25–26
matrices, 21–22
numerics, 25
sparse matrices, 28–29
tables, 30
tall arrays, 28–29
MATLAB toolbox, xiv
Membership functions
bell function, 161
Gaussian, 161
sigmoidal, 161
trapezoid, 161
triangular, 161
MHTAutomobileDemo function, 375
MHTBilliardsDemo, 342–343
Mixed-integer linear program (MILP), 422
Model Reference Adaptive Control (MRAC)
adaptation parameters, 137
closed-loop system, 136
differential equations, 139
disturbance angular acceleration, 136
function, 139–140
gain convergence, 144
implementation, 138
MIT rule, 136
parameters, 137
problem, 136
for rotor, 136, 142–143
RotorSim, 140
RungeKutta, 139
solution, 136
Models, 2, 360, 364, 368, 381, 415
multiple model filters, 331
Monte-Carlo methods, 95
Multilayer feedforward (MLFF) neural
networks, 15, 240, 263
Multiple hypothesis testing (MHT), 355, 374
code listing, 353
GLPKMEX program, 329
terms, 328
track management software, 329
track-oriented approach (see
Track-oriented MHT)
N
Neural aircraft control
activation function, 202–203
enumeration, all sets of inputs, 208–210
functions and scripts, 228
longitudinal dynamics, 191, 192
longitudinal motion, 191–198
neural net
for learning control, 204–208
for pitch dynamics, 223–226
sigma-pi, 210–213
441INDEX
nonlinear simulation, 226–228
numerical simulation, 200–202
numerically finding equilibrium,
198–200
PID control
implement, 213–218
of pitch, 218–223
Neural net designing, pendulum
neural estimated angles, different
magnitude oscillation, 243, 244
neural net results, 243
NeuralNetMLFF, 240, 242
NeuralNetTraining, 241
NNPendulumDemo, 240
training data structure, 242
training error, 242
Neural net functions, creation
backpropagation, 266
identify digits, 262
NeuralNetMLFF, 265
NeuralNetTraining function, 266
neuron activation functions, 263
Neuron function, 263, 264
sigmoid logistic function, 263, 264
Neural Net Trainer GUI, 269
Neural networks, 14, 388, 404, 411, 420
cat images, 320
CNN (see Convolutional neural network
(CNN))
deep learning neural net, 303, 304
definition, 14
general neural net, 303
image processing, 318
for learning control
aircraft control system, 204
pinv function, 206
pitch angular acceleration, 204
recursive training/learning, 207–208
RecursiveLearning, 206–207
sigma-pi, 204
for pitch dynamics, 223–226
sigma-pi (see Sigma-pi neural network)
NeuralNetDeveloper tool, 262, 267
NeuralNetMLFF, 240, 242, 265, 266
NeuralNetTrainer, 271
NeuralNetTraining function, 266
Neurons, 264, 265
NNPendulumDemo, 240
Non-fuzzy bang-bang controller, 187
Nonlinear measurement, 9
Nonlinear simulation
acceleration magnitude and angles, 226,
227
addLearning, 226
aircraft accelerations, 226, 228
aircraft pitch angle change, 226, 227
AircraftSim, 226
Nonzero elements, 29
Numerical simulation of aircraft
AircraftSimOpenLoop, 201
open-loop response to pulse, 202
RHSAircraft, 200
RungeKutta, 201
Numerically finding equilibrium
CostFun, 199–200
EquilibriumState, 199–200
fminsearch, 200
forces and torques balance, 198
Jacobian, 199
vertical velocity and thrust, 200
O
Object recognition, 420
oneDigitMode, 261
OneNeuron, 237, 238
Online learning, 4
Optimal control systems, 7
Optimization, 6, 419, 423
CVX, 425
GLPK, 424
LOQO, 424
SeDuMi, 425
SNOPT, 424
YALMIP, 425
442INDEX
P, Q
Parallel Computing Toolbox, 36
Parameter pairs, 49
Passing control
algorithms, 362
AutomobilePassing function, 362–363
maneuvers, 362
passState variable, 362
Pattern recognition, 7–8
PD controller/proportional derivative, 214
PDF, see Probability density function (PDF)
Pendulum, 231
dynamics, 231
linear and nonlinear equations, 235
PendulumSim, 234
results, models, 235
rigid body rotation, 231
rigid connection, 231
RungeKutta function, MATLAB
integrator, 233
solution, 231
torque, 232
PinholeCamera function, 399
pinv function, 206
Pitch angle error, 213
Pitch dynamics, 191, 192, 223–226
PitchDynamicInversion, 224, 226
PitchNeuralNetTraining, 224–225
Princeton Plasma Physics Laboratory (PPPL),
xiii
Princeton Satellite Systems, 341, 361
Core Control toolbox, 421
target tracking, 422
PrintFuzzyRules, 169
Probabilistic graphical model, 18
Probability density function (PDF), 86
Proportional integral differential (PID)
controller
implement
closed-loop transfer function, 215
code for, 215–218
derivative operator, 214
double integrator equations, 213, 217
feedback controller, 214
PD controller/proportional
derivative, 214
pitch angle error, 213
unit input, 217
and neural network, 204
of pitch
acceleration magnitude
in aircraft pitch and elevator angle,
221, 222
PID vs. DI, 221, 223
aircraft lift and drag, 221, 222
aircraft states during pitch angle change,
221
AircraftSim, 219–221
dynamic inversion function, 218–219
elevator movement, 218
simulation setup, 219–221
R
Radar systems, 327, 330, 355
RandomYawAngles function, 399
Recursive training/learning, 207–208
RecursiveLearning, 206–207
Regression, 10–13, 391, 417, 420, 423
Reinforcement learning, 418
RHSAircraft, 195–196, 200
RHSAutomobile function, 375
RHSAutomobileXY function, 375
RMS error, 270
Robocalling systems, 17
RungeKutta function, 201, 233
S
SaveTS, 260
Scikit-learn, 423
Security system, 5
SeDuMi, 425
Self-tuning
damping ratio, 132
equations, 129
443INDEX
Fast Fourier Transform, 128, 131
frequency, 130
frequency spectrum, 132–133
parameter identification, 128
RHSOscillator dynamical model, 130
Spring-mass-damper system, 129
TuningSim calls FFTEnergy, 133
Semi-supervised learning, 4
Set theory framework, 157
Ship steering
dynamical equations, 145
nonlinear problem, 145
parameters, 146
quadratic regulator generator code, 147
simulation, 146, 148, 149
Sigma-pi neural network, 204, 205, 208, 210,
223
actions, 210
data structure format, 210
default data structure, 211
kSigmoid, 211
recursive learning/training, 212–213
SigmaPiNeuralNet, 210–212
switch statement, 210
truth model, 212
SigmaPiNeuralNet, 210–212, 226
Sigmoidal membership function, 161
Sigmoid function, 203, 267
Sigmoid logistic function, 263, 264
Sigmoid magnitude function, 273
SimpleClassifierDemo, 295–298
Single-digit network, 271, 272
Single-digit training error, 270
Single neuron angle estimator
activation functions, 236, 237
linear pendulum equation, 237
neuron outputs, 239
one-neuron function, linear activation
function, 239
OneNeuron, 237, 238
pendulum dynamics comparison, linear and
tanh neuron output, 239
problem, 235
real neuron, 236
solution, 236
two-input neuron, 236
Smart wiper control system, 169–170
SmartWipers, 160–161, 169
SmartWipersDemo, 169
SmartWipersSystem, 169
SmartWipersTest, 172–174
SNOPT, 424
Spacecraft
attitude determination
AttitudeDetermination
function, 391
averagePooling2dLayer, 393
batchNormalizationLayer, 392
CNN, 391, 392
convolution2dLayer, 392
deep learning, 391
dropoutLayer, 394
fullyConnectedLayer, 393
maxPooling2dLayer, 393
regressionLayer, 394
reluLayer, 392–393
scatter plot, 397
star images, 395
star pattern, 377
testing, 397–398
training and validation sets, 395
training interface, 397
training setup, 395–396
yaw angle histogram, 395, 396
YawToImages, 394
autonomous control, 416
boolean logic, 402
camera model (see Camera model)
camera views, attitude simulation
AttitudeSim function, 384, 386–387
conventions, 385
notation, 385
quaternions, 384–386
rotation, 386
444INDEX
scalar/vector components, 384, 385
transformation, 385
celestial sphere, 383–384
jet select logic, 401–402
rules, 402
Space Shuttle Orbiter thruster locations,
402, 403
star camera, 377
thrusters, 401, 402
yaw angle rotation, 387–388
yaw images
grayscale, 388
Label.mat file, 388–390
transformed pixels/figure, 388
YawPixelTransform function,
388–390
YawToImages function, 388, 390
Spacecraft pointing
estimated and actual inertia, 152
spacecraft model, 148, 150
states and control outputs, 151, 152
Sparse matrices, 28–29
Spring-mass-damper system
acceleration, 89
damping ratios, 89, 90
default data structure, 91
differential equations modeling, 89
dynamical equations, 90
simulation, 91–92
Square wave input, 140–143
Star catalog
ECI, 379, 380
Greenwich, 380
Hipparcos, 378, 380
LoadHipparcos function, 378–379
pinhole camera, 381
precession and nutation, 381
sprintf, 379
star image, 378
State estimator, linear Kalman filter, 87
angle measurement geometry, 87, 88
solution, 88
spring-mass-damper system, 87, 88
Statistical methods, 420
Supervised learning, 3
Support vector machines (SVMs), 16, 417,
420, 423
switch statement, 210
Synonym set, 305
System identification, 5, 7, 416, 419
T
Table Data, 38–41
tanh activation function, 238
Taxonomy, 6–8, 355, 401, 420
Technology readiness levels (TRLs), 157
Testing, neural network
DigitTrainingData, 272
neural net results with sigmoid/step
activation functions, 273
Time-sharing, 412
Titan landing control systems, 16
Track-oriented MHT, 18
algorithm, 331–333
billiard ball
Kalman Filter, 335–342
MHT, 342–345
covariance, 331
hypothesis formation, 334–335
measurement assignment and tracks,
333–334
measurements, 330, 331
one-dimensional
MHT, 349–351
motion, 345–349
track pruning, 335
tracking problem, 329, 330
uncertainty ellipsoids, 329–330
Track pruning, 329, 335, 336, 422
Traditional machine learning, 6
Training
data, 404
online, 4, 305
semi-supervised, 4
445INDEX
sets, 3
supervised, 3
unsupervised, 3
Training, neural network
with multiple outputs
multiple-digit neural net weights, 276,
277
NeuralNetMLFF, 276
training RMS, 274, 275
with one output node
DigitTrainingTS, 269
inputs, 268
layer 2 node weights and evolution,
biases, 270
Neural Net Trainer GUI, 269
NeuralNetTrainer, 271
RMS error, 270
single-digit training error, 270
10 node hidden layer weights, 272
30 node hidden layer weights, 270, 271
Trapezoid membership function, 161
Triangular membership function, 161
TRLs, see Technology readiness levels (TRLs)
Tuning simulation results, 135
TurboSquid, 364
Two-dimensional line graphs (2D line plots),
49–52
U
UKFAutomobileDemo function, 375
Unscented Kalman Filter (UKF), 9, 372
for parameter estimation
covariance, 118
outcomes, 122
problem, 117
RHSOscillator, 120
RHSOscillatorUKF, 121
solution, 118
UKFPSim, 119
UKFPWeight, 121
for state estimation
Cholesky factorization, 116
Gaussian processes, 112
KFInitialize, 116
measurements, 113
nonlinear, 111
outcomes, 117, 118
solution, 112
UKFPredict, 114–115
UKFWeight, 113
Unscented Kalman Filters (UKF), 421
Unsupervised learning, 3
V
Variable HVAC fuzzy controller
AC/heat setting, 188
demo function, 181
inputs and outputs, 181, 186
mode output, 187
mode value, 184
non-fuzzy bang-bang controller, 187
sign function, 184
simulation, 187–189
string values, 181
W
Watson, 414
Window wiper fuzzy controller
BuildFuzzySystem, 169
data structure, 169
fuzzy inference, 169
FuzzyPlot, 169
FuzzyRand, 172
printed rules, 170–171
PrintFuzzyRules, 169
rotate3d, 171–172
smart wiper control system, 169–170
SmartWipersDemo, 169
SmartWipersSystem, 169
SmartWipersTest, 172–174
X
XOR
data division, 245
deep learning network GUI, 246
446INDEX
two hidden layers, 251, 252
DLXOR.m script, 244
GUI progress, 246–247
MATLAB GUI, 245
network training
histogram, 247, 249
performance, 247
state, 247, 248
performance, 245
plots, 247
regression, 247, 250
training, 245
Y, Z
YALMIP, 425
YawPixelTransform function, 387–388, 399
YawToImages function, 388, 394, 399
YawToPixelsDemo function, 399
#ماتلاب,#متلاب,#Matlab,#مات_لاب,#مت_لاب,
كلمة سر فك الضغط : books-world.net
The Unzip Password : books-world.net
أتمنى أن تستفيدوا من محتوى الموضوع وأن ينال إعجابكم
رابط من موقع عالم الكتب لتنزيل كتاب MATLAB Machine Learning Recipes - A Problem-Solution Approach
رابط مباشر لتنزيل كتاب MATLAB Machine Learning Recipes - A Problem-Solution Approach 
|
|
