An Independent Study course is one that is designed by the student to pursue in-depth study in a particular area of interest. An Independent Study can be taken for 9 or 12 units. Each Independent Study, regardless of the proposed units, fulfills the requirement of one elective course only. A maximum of two Independent Studies are permitted to apply toward degree completion requirements.
The purpose of the Independent Study is for the student to further explore an area of interest that he or she may have just scratched the surface of while taking one of the core courses.
While the Independent Study is intended for individual students, it may be completed in very small teams. All members of the team must have completed the core courses in the MSE/MSIT program. A team project must be more significant in scope than an individual effort.
Note: All Independent Study proposals are due no less than 2 weeks prior to the completion of the semester preceding the semester in which the proposed Independent Study course will take place.
The steps for the Independent Study are as follows:
Propose a project. This will take the form of a formal document and will be considered the first deliverable of the project. The proposal must include:
an executive summary
a definition of the work to be completed, including
an advisor (a Carnegie Mellon faculty member)
a proposal of how you want the grade to be generated from the independent material
Deliverables will always include a report that summarizes the strengths and weaknesses of the technique in the chosen domain and relate what went well and poorly during the Independent Study.
Discuss the project.
Begin the project.
Report periodically to your faculty mentor.
Independent Study Samples
Combinatorial Testing (2012)
This project assessed the suitability of a method proposed by NIST called “combinatorial coverage” to the evaluation of a test suite used to verify a software component using the Executive Services (ES) subsystem of the Core Flight Executive (cFE) system from NASA. The study assessed the adequacy of a sample of the test program that was packaged in the open source distribution of the cFE framework with regards to the degree of interaction or combinatorial coverage present in its test suite. As evaluated, this technique was shown to be a useful measure for the quality of a software test program design, before effort is expended to implement the test procedures and test scripts, which would be a valuable endeavor for a software organization with limited budgets and schedule for software testing.
Formal Probabilistic Modeling and Model Checking (2009)
To research the PRISM probabilistic model checker, applying it to a real world scenario. Many current model checking applications do not allow the user to specify probabilities for non-deterministic system behavior. Instead they randomly select a path with an equal probability of reaching any available state. PRISM’s ability to specify the probability of arriving at a given state appears to have significant real world applications.