Preview Workflow

The CIM Courses system will be down temporarily undergoing routine maintenance.

Viewing: BME 6850 : Pattern Recognition

Last approved: Fri, 03 Jun 2016 04:23:29 GMT

Last edit: Thu, 02 Jun 2016 15:20:45 GMT

Catalog Pages referencing this course
School of Engineering and Applied Sciences
Biomedical Engineering (BME)
BME
6850
Pattern Recognition
Pattern Recognition
Fall 2016
3
Course Type
Lecture
Default Grading Method
Letter Grade

No
Yes
ECE 6850 - Pattern Recognition
ECE 6015
Corequisites

25

Frequency of Offering
Every Year
Term(s) Offered
Fall
Spring
Summer
Are there Course Equivalents?
Yes
 
ECE 6850 - Pattern Recognition
No
Fee Type


No


Random vectors, transformations. Hypothesis testing, error probability: bias, variance, and sample size, resampling; sequential methods. Bayes, other linear classifiers. Discriminant functions, support vector machines, maximum-likelihood and parameter estimation, dimensionality reduction. Nonparametric methods; unsupervised learning and clustering; feature selection and ordering. Applications in industry and medicine. Student projects. Learning is reinforced by homework problems and in-class and at-home computer examples.
1. Demonstrate an understanding of the methods of data representation and classification in high-dimensional spaces, the intrinsic dimensionality and variability of data, and the importance of testing and validation. 2. Learn techniques of decision theory, cost and risk, and error bounds; develop competence in applying to real data 3. Demonstrate an understanding of parameter estimation and of nonparametric techniques 4. Learn principles and implementations of linear discriminant functions, perceptrons, and support vector machines; apply to real data 5. Demonstrate an understanding of machine learning, resampling methods, and combination classifiers 6. Demonstrate an understanding of unsupervised learning and clustering

Course Attribute


Key: 9837