Course
of Peripheral Equipment and Human Computer Interface |
Teacher:: PhD, Associate Professor Dobrea Dan Marius |
E-mail: mdobrea@etti.tuiasi.ro |
Year of study:
Specialization:
Department:
Faculty:
University:
Office:
Home page:
Laboratory facilities: |
4th
Computer engineering
Department of Applied Electronics and Intelligent Systems
Faculty of Electronics
and Telecommunications
Tehnical University "Gh.
Asachi"
room III-26
http://www.etc.tuiasi.ro/cin/Members/Dan/dan.htm
http://www.etc.tuiasi.ro/cin/Courses/Equipments.htm |
Course Description:
Human-Computer Interface is, basically,
a discipline concerned with the design, evaluation, and implementation of
interactive computing systems for human use; its aim is to build "intelligent"
programs or systems that are sensitive to the need of the user through the
physical and psychological state identification.
This course and its labs present techniques for external data acquisition
and introduce techniques for pattern recognition in order to classify the
extracted physiological signals features.
Simultaneously with the course interactive demonstrations, in NeuroSolution
and Matlab, are presented. These interactive demonstrations intend to stimulate
interest and help students to gain intuition about how classifier systems
work under a variety of situations and constrains.
Upon completion of this curse, students should be able to: acquire biomedical
signals, have fundamental background in classifier system, be able to implement
a classifier system in C++/C language and understand the role and drawbacks
of each stage and method used in pattern recognition systems.
|
|
Class Meets:
Each Wednesday, starting at 14:00 and goes up to 17:00 (with
10 minutes break at 14.50 and 15.50) in Room P2, from February 17 to
16 May
Back |
Examination:
Final Exam: 60% (The final exam will be closed book and
will cover material from lectures and homework assignments.)
Project: 20%
Homework: 10%
Class Participation: 10%
Back
|
The course cover the following topics:
1. Introduction
2. Cognitive limitation and particularity of human acquisition
process
3. Phase stages in pattern recognition
3.1. Data acquisition
3.2. Signal preprocessing/processing
3.3. Features generation
3.4. Clustering
3.5. Classification
3.5.1. Introduction
3.5.2. Decision surface
3.5.3. Discriminate functions
3.5.3. Metrics
4. Elementary classifiers
4.1. Template
matching
4.2. Minimum-distance classifiers
4.3. Limitations of simple classifiers
4.3.1. The features
may be inadequate
4.3.2. The features may be correlated
4.3.3. The decision boundaries are not anymore linear
4.3.4. There are distinct subclasses in the data
4.3.5. The feature space may simply be too complex
5. Statistical Classifiers
5.1. Statistical
Overview
5.1.1. Random vectors and their characterisation
5.1.2. Expectation and moments
5.1.3. The Gaussian density function
5.1.4. Linear transformation of random vectors
5.1.5. Diagonalization by unitary transformation
5.1.6. Diagonalization by triangular decomposition
5.2. Mahalanobis Classifiers
5.2.1. Mean and
Variance
5.2.2. Mahalanobis Metric. Mahalanobis
Classifiers.
5.3. Bayesian Classifiers
5.3.1. Optimal
Decision Boundary Based on Statistical Models of Data
5.3.2. A Two-Dimensional Pattern-Recognition Example
5.3.3. Discriminant Sensitivity to the Size of the Data
5.3.4. Features Selecton Baze on PDF
5.4 Nonparametric
Method Used for Estimation and Classification
5.4.1. Density
Estimation
5.4.2. Parzen Windows. Parzen Estimation.
5.4.3. K Nearest Neighbor Density Estimation
5.4.4. Nearest Neighbour Classification
5.4.5. The Nearest Neighbour Rule for Two Classes and N Classes
5.4.6. Error Rate fot the Nearest Neighbour Rule
6. Artificial Neuronal Network
6.1. Introduction
6.2. The Perceptron
6.2.1.
Pattern Recognition Ability of the McCulloch-Pitts PE
6.2.2. The Perceptron
6.3. One Hidden Layer Multilayer
Perceptrons
6.3.1.
Discriminant Functions
6.3.2. Training
the One Hidden Layer MLP
6.3.3. The Effect of the Number of Hidden Neurons
6.4. MLPs with
Two Hidden Layers
6.4.1. Discriminant
Functions
6.4.2. MLPs as Universal Classifier
6.5. Designing and Training MLPs
6.5.1. Learning
Process Control
6.5.2. Methods to Improve of the Learning Process
6.5.3. Stop Criteria
6.5.5. Error Criterion
6.5.6. Network Size and Generalization
Back
|
Lab syllabus:
- The graphical user interface of
the LabWindowsCVI (2 hours)
- How to use timers in LabWindowsCVI ?
(2 hours)
- The Graphic Controls of the LabWindowsCVI
developement environment. (2 hours)
- The advance analysis library of the
LabWindowsCVI (2 hours)
- The parallel port (LPT) (2 hours)
- The serial port 1 - Analog to digital
conversion (2 hours)
- The serial port 2 - LCD communication
(micrcontroler) (2 hours)
- National Instrument - Digital Acquisition
Board (2 hours)
- A noncontact system used to respiration
acquisition (2 hours)
- Visual pattern recognition (2 hours)
- K-means clustering and the silhouette
parameter (2 hours)
- Introduction into the NeurSolution environment
for pattern classification (2 hours)
- Brain Computer Interface 1 - EEG preprocessing
and features extraction (2 hours)
- Brain Computer Interface 2 - EEG classification
(2 hours)
Back
|
Bibliography:
- Charles W. Therrien, Discrete
Random Signals and Statisitcal Signal Processing, Printice-Hall
International Inc., New Jersey, United States of America, 1992, ISBN 0-13-217985-7
- José C. Principe, Neil R. Euliano, W. Curt
Lefebvre, Neural and Adaptive Systems: Fundamentals
Through Simulations, John Wiley & Sons Inc., United States
of America, ISBN: 0-471-35167-9, 2000
- Liviu Goras, Semnale
Circuite si Sisteme, Editura "Gh. Asachi", Iasi,
Romania, 1994, ISBN 973-96222-8-3
- Richard O. Duda, Peter E. Hart, David G.
Stork, Pattern Classification,
John Wiley & Sons Inc., New York, United States of America, 2001, ISBN
0-471-05669-3
- Victor Neagoe, Octavian Stanasila, Teoria
Recunoasterii Formelor, Editura Academiei Romane, Bucuresti,
România, 1992, ISBN: 973-27-0341-5
Back |
Additional Literature:
- R. Picard, E. Vyzas, J. Healy, Toward
Machine Emotional Intelligence: Analysis of Affective Physiological State,
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 23,
no. 10, pp. 1175 – 1191, 2001
- K. H. Kim, S. W. Bang, S. R. Kim, Emotion
recognition system using short-term monitoring of physiological signals,
Medical & Biological Engineering & Computing 2004, Vol. 42, pp.
419 – 427, 2004
Back
|
Student zone
Back |