Expected learning outcomes
- Candidates should have a solid grounding in core concepts of forensic intelligence, forensic data analytics and their application to large-scale, big data investigations.
- Candidates possess advanced knowledge of forensic analytics methods, their functional requirements and constrains.
- Candidates have advanced knowledge of common methods for search, identification, verification/matching and clustering of forensic data.
- Candidates have thorough knowledge on the theory and methods underlying machine learning, pattern recognition and data mining.
- Candidates are capable of applying relevant methods for independent analysis and mining of forensic data, e.g. binaries, text, time signals, images and video.
- Candidates are able to analyze and critically review literature in the field of applied pattern recognition and data mining, and are able to apply results from the literature in structuring and formulating arguments and reasoning on forensic intelligence topics.
- Candidates are able to plan and conduct a limited, guided case project based on primary literature resulting in a reasoned and coherent report.
- Candidates are able to adopt knowledge and methods in the area of forensic intelligence to novel fields so as to be able to successfully complete advanced tasks and projects in forensic intelligence and data analytics.
- Candidates are able to work independently and are familiar with core concepts and problems in forensic intelligence and data analytics.
- Candidates are able to contribute to innovations and innovative processes, identifying advanced forensic intelligence problems, and computational methods contributing to their solution.
- Introduction to machine learning, pattern recognition and data mining
- Knowledge and evidence representations with attributes, and attribute quality measures
- Forensically sound methods for evidence filtering, search, data mining, pattern recognition, and link analysis
- Measure of method-performance evaluation and reliability
- Automation of logfile analysis, file carving, evidence search, crime detection and prevention
- Applications in forensic multimedia-content analysis, e.g. video, images, audio
- Knowledge extraction, automatic categorization, and indexing of multimedia content
- Content analysis of multimedia data in the absents of structure and metadata
- Detect and reconstruct illegal activities from multimedia content, and use it as a source of intelligence
Teaching Methods (additional text)
Other (Independent study)
Other (Essay/Article writing)
The course will be made accessible for both campus and remote students. Every student is free to choose the pedagogic arrangement form that is best fitted for her/his own requirement. The lectures and exercises in the course will be given on campus and are recorded, so that the course is open for both, campus and remote students. Participation in the exercises is recommend since they aim at deepening the lecture topics and prepare for the assignments that will be graded. Project work in the second half of the course will promote the practical application of data analytics to realistic forensic case scenarios. All the lecture material will be available on Internet through GUC’s learning management system (Fronter).
Form(s) of Assessment
Form(s) of Assessment (additional text)
Assessment consists of three parts, pass decision is on cumulative grade of both parts:
- Part 1 is a written examination (3 hours), accounting for 40% of grade
- Part 2 are four assignments, accounting for (4x5%) of grade
- Part 3 is an assessment of the essay/article with up to 2000 words (40%)
The essay/article is evaluated by an internal examiner.
Alphabetical Scale, A(best) – F (fail)
Evaluated by external and internal examiner.
Re-sit August 2016 for the Written exam.
New assignment(s) must be provided if one or several of these are failed.
D: No printed or hand-written support material is allowed. A specific basic calculator is allowed.
The following textbook is the primary references. Additional sources, e.g. presentation material will be provided during the course.
- Igor Kononenko, Matjaz Kukar (2007). Machine Learning and Data Mining, Woodhead Publishing, ISBN-10: 1904275214, ISBN-13: 978-1904275213
This course is based on and overlapping the existing IMT4612 Machine Learning and Pattern Recognition 1, and IMT4641 Computational Forensic.