University of Waterloo COVID-19 update

Please see the University of Waterloo’s frequently asked questions for information about COVID-19 and how it has affected university operations.

Please visit the list of modified services if you have questions about university services.

Although the University of Waterloo is closed for in-person events until further notice, many virtual events and presentations of interest to computer scientists are taking place each week at the Cheriton School of Computer Science. Please check out what our students and faculty are doing.

PhD Seminar • Software Engineering — Transfer Learning for Improving Model Predictions in Highly Configurable SoftwareExport this event to calendar

Tuesday, July 24, 2018 — 11:00 AM EDT

Pavel Valov, PhD candidate
David R. Cheriton School of Computer Science

Modern software systems are built to be used in dynamic environments using configuration capabilities to adapt to changes and external uncertainties. In a self-adaptation context, we are often interested in reasoning about the performance of the systems under different configurations. Usually, we learn a black-box model based on real measurements to predict the performance of the system given a specific configuration. However, as modern systems become more complex, there are many configuration parameters that may interact and we end up learning an exponentially large configuration space. Naturally, this does not scale when relying on real measurements in the actual changing environment. 

We propose a different solution: Instead of taking the measurements from the real system, we learn the model using samples from other sources, such as simulators that approximate performance of the real system at low cost. We define a cost model that transforms the traditional view of model learning into a multi-objective problem that not only takes into account model accuracy but also measurements effort as well. We evaluate our cost-aware transfer learning solution using real-world configurable software including (i) a robotic system, (ii) three different stream processing applications, and (iii) a NoSQL database system. The experimental results demonstrate that our approach can achieve (a) a high prediction accuracy, as well as (b) a high model reliability.

Location 
DC - William G. Davis Computer Research Centre
2314
200 University Avenue West

Waterloo, ON N2L 3G1
Canada

S M T W T F S
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1
2
3
4
  1. 2020 (109)
    1. June (8)
    2. May (17)
    3. April (20)
    4. March (17)
    5. February (25)
    6. January (22)
  2. 2019 (255)
    1. December (21)
    2. November (25)
    3. October (16)
    4. September (20)
    5. August (18)
    6. July (12)
    7. June (23)
    8. May (23)
    9. April (32)
    10. March (25)
    11. February (16)
    12. January (24)
  3. 2018 (220)
  4. 2017 (36)
  5. 2016 (21)
  6. 2015 (36)
  7. 2014 (33)
  8. 2013 (23)
  9. 2012 (4)
  10. 2011 (1)
  11. 2010 (1)
  12. 2009 (1)
  13. 2008 (1)