MATH3067 Information and Coding Theory

This page contains information on the Senior normal Unit of Study MATH3067: Information and Coding Theory.

• Taught in Semester 2.
• Credit point value: 6.
• Classes per week: Three lectures and one tutorial.
• Lecturer(s): Bob Howlett and Claus Fieker.

Please refer to the Senior Mathematics and Statistics Handbook for all questions relating to Senior Mathematics and Statistics. In particular, see the MATH3067 handbook entry for further information relating to MATH3067.

You may also view the Faculty Handbook entry for MATH3067 from the central units of study database.

The first six weeks will be on Information Theory; the lecturer will be Bob Howlett. The remaining seven weeks will be on Coding Theory; the lecturer will be Claus Fieker.

Assessment Information

There will be one assignment for the Information Theory half of the unit and one assignment for the Coding Theory half. These two assignments will each carry 5% of the total assessment for the unit. There will also be a tutorial participation mark, worth 10% of the total assessment. The remaining 80% of the assessment will come from a two-hour exam at the end of the semester. (The exam will be half on Information Theory and half on Coding Theory.)

The Information Theory assignment questions will be released at the end of the 5th week, and the assignment should be handed in at the lecture on the Monday of Week 7, or earlier. Details of the Coding Theory assignment will be announced later.

Textbooks

For the Information Theory half students should obtain the notes "Mathematics 3067 Information Theory" by N. R. O'Brian. These will be made available for purchase from KopyStop.

This book is now available from KopyStop for $14.50.

Copies of the notes for individual lectures will be released here progressively. These will usually be scanned copies of the lecturer's handwritten notes.

Correction: In Lecture 11 I gave a non-standard definition, which should be changed. I said that a Markov process is called stationary if the transition probabilities are constant (independent of time). The correct term is time-homogeneous. It should not be called stationary unless an additional requirement is satisfied, namely that probability distribution for the states is also constant. In other words, the initial distribution should be the stationary distribution for the given probability transition matrix.

Tutorial sheets and solutions will be released here progressively. Note that tutorials do not start until Week 2.

• Tutorial 1 and Tutorial 1 Solutions
• Tutorial 2 and Tutorial 2 Solutions
• Tutorial 3 and Tutorial 3 Solutions
• Tutorial 4 and Tutorial 4 Solutions

From Monday 11th August (inclusive) the Monday lecture will be in Carslaw 175 instead of Carslaw 350