Login for PhD students/staff at UCPH      Login for others
Machine Learning and Imaging Methods (MLI-M)
Provider: Faculty of Science

Activity no.: 5173-19-02-32 
Enrollment deadline: 15/04/2019
PlaceDIKU
University of Copenhagen, 2300 KĂžbenhavn S
Date and timeMay 2019 - June 2019
[antalgange]5
Regular seats60
Activity Prices:
  - Deltager/Participant from SCIENCE0.00 kr.
  - Deltager/Participant Others3,600.00 kr.
ECTS credits3.00
Contact personAnnette Lange    E-mail address: alange@di.ku.dk
Enrolment Handling/Course OrganiserAnnette Lange    E-mail address: alange@di.ku.dk
Written languageEnglish
Teaching languageEnglish
Semester/BlockBlock 4
Scheme group notePlenum teaching will be done on selected Wednesdays
Exam formWritten assignment
Exam detailsThe students need to hand in their report (shortly after the final course day). The report must be approved. The students are allowed to work in 2-person groups.
Grading scalePassed / Not passed
Exam re-examinationIf the student's synopsis is not approved, the student has teh possibility of resubmission based on the feedback from the examination.
Course workload
Course workload categoryHours
Course Preparation5.00
Lectures15.00
Theory exercises35.00
Project work15.00
Exam1.00

Sum71.00


Content

IMPORTANT ABOUT SIGNING UP!
Please sign up via this link- as Frontplanner is not cooperating correctly

https://datalab.science.ku.dk/english/course/phd-courses/mli-methods/

 

The Machine Learning and Imaging Methods (MLI-M) course introduces key analysis methods in Machine Learning and Image Analysis. These allow investigations of scientific data from most fields, including data from physical measurements, questionnaires, pictures, Internet searches, and biochemical outcomes. We cover data cleaning (e.g. missing data, denoising), feature extraction, machine learning basics (labels, variables, parameter optimization, overfitting, cross-validation), key machine learning and image analysis methods based on both unsupervised and supervised learning, and visualization. Method-wise, we start at Linear Discriminant Analysis and end with Deep Learning.
At the end of the course, the students must write a synopsis with a suggestion for an analysis ideally performed on their own data. This synopsis could form the basis for the MLI Projects PhD course offered by the Data Science Lab in September 2019


Aim and content

The Machine Learning and Imaging Methods (MLI-M) course introduces key analysis methods in Machine Learning and Image Analysis. These allow investigations of scientific data from most fields, including data from physical measurements, questionnaires, pictures, Internet searches, and biochemical outcomes. We cover data cleaning (e.g. missing data, denoising), feature extraction, machine learning basics (labels, variables, parameter optimization, overfitting, cross-validation), key machine learning and image analysis methods based on both unsupervised and supervised learning, and visualization. Method-wise, we start at Linear Discriminant Analysis and end with Deep Learning.

At the end of the course, the students must write a synopsis with a suggestion for an analysis ideally performed on their own data. This synopsis could form the basis for the MLI Projects PhD course offered by the Data Science Lab in September 2019.


Formel requirements


The number of participants is limited at 60, and priority will be given to students who previously followed another Data Science Lab course (Introduction to Python or R, Statistical Methods I).
We assume that the students have some experience with Python programming.

Learning outcome
After course completion the students are expected to be able to:

Knowledge:
- Understand key machine learning concepts (parameter training, overfitting).
- Understand key machine learning methods (LDA, (un-) supervised learning).
- Understand key image analysis methods (feature detection).
Skills:
- Develop/adapt/extend a computer-based software method for quantification and/or analysis of relevant data.
Competences:
- Propose relevant analysis methods for scientific data science problems.
- Consider cross-disciplinary data science methods in their research.

Literature
Course lecture slides and exercises. We will use data, examples, and other material from publicly available online sources.

Teaching and learning methods
The course is composed of sessions combining lectures and exercises. For each topic, the students will get hands-on experience in applying, modifying, and programming analysis methods.
We assume that the students have some experience with Python programming.

Lecturers
Erik Dam and Raghavendra Selvan

Workload
Course workload category Hours
Course Preparation 5.00
Lectures 15.00
Theory exercises 35.00
Project work 15.00
Exam 1.00
________________________________________
Sum 71.00
________________________________________

Search
Click the search button to search Courses.


Course calendar
See which courses you can attend and when
JanFebMarApr
MayJunJulAug
SepOctNovDec



Publication of new courses
All planned PhD courses at the PhD School are visible in the course catalogue. Courses are published regularly.