Kumara Sastry » Interactive Evolutionary Algorithms

Speeding online synthesis via enforced selecto-recombination

Kumara Sastry — Sun, 06 Apr 2008 05:32:39 +0000

Saruwatari, S., Llorà, X., Yasui, N. I., Tamura, H., Sastry, K., Goldberg, D. E. (2008). IlliGAL Report No. 2008004. University of Illinois at Urbana-Champaign, Urbana IL. [Full Paper - PDF] [Full Paper - PS].

Abstract:
Brainstorming has been greatly used as a method to generate a large number of ideas by variety of each participant’s knowledge. However, brainstorming does not always work well because of spatial, communication limitations. Moreover, brainstorming techniques present limited scalability. Meanwhile, genetics algorithms have been mostly regarded as an engineering or technological tool. However, the innovation intuition suggests that genetic algorithms may be also regarded as models of human innovation and creativity. This paper focuses on online creativity sessions. Modeling those creative efforts using selecto-recombinative mechanism can provide three times more novel ideas, increase the posting frequency by a 2.6 factor, and help overcome superficiality on online communications by favoring synthetic thinking.

Discovering building blocks for human based genetic algorithms

Kumara Sastry — Sun, 06 Apr 2008 04:01:19 +0000

Ueda, T., Yasui, N. I., Llorà, X., Sastry, K. Goldberg, D. E. (2008). Smart Systems Engineering: Computational Intelligence in Architecting Complex Engineering Systems (ANNIE 2007). [First Runner Up, Theoretical Developments in Computational Intelligence] [Preprint: IlliGAL Report No. 2007020].

Abstract:
The push for rapid innovation and creativity in this Internet age places a premium on eective integration of both human and computer-generated knowledge. One of the key components of a distributed and scalable environment for accomplishing this integration called DISCUS is the human-based genetic algorithm (HBGA)–a GA where humans perform genetic operations. This paper takes the first step towards designing a competent HBGA, which can enable humans to innovate quickly, reliably, and accurately. Specifically, this paper proposes a methodology for discovering building blocks from text documents including reports, chat, transcripts and e-mail. The proposed method has been applied to simple test problems and to a news article set. The results show that the proposed BB-identification methodology is eective and enables humans to eectively exchange the BBs for rapid innovation.

Analyzing active interactive genetic algorithms using visual analytics

Kumara Sastry — Fri, 14 Jul 2006 12:11:59 +0000

Llorà, X., Sastry, K., Alías, F., Goldberg, D. E., Welge, M. (2006). Proceedings of the 2006 Genetic and Evolutionary Computation Conference, 1417—1418. [Full paper - PDF] [Full paper - PS].

This paper build on active interactive genetic algorithms and introduces visual-analytic techniques to aggregate, summarize, and visualize the information generated during interactive evolutionary processes. Special visualizations of the user-provided partial ordering of solutions, the synthetic fitness surrogates induced, and the model of user preferences were prepared. The visual- analytic techniques proposed point out potential pitfalls, strengths, and possible improvements in a non-trivial case study where the hierarchical tournament selection scheme of an active interactive genetic algorithm is replaced by an equivalent incremental selection scheme. Visual analytics provided an intuitive reasoning environment that unveiled important properties that greatly affect the performance of active interactive genetic algorithms that could not have been easily reveled otherwise.

Efficient interactive weight tuning for TTS synthesis: Reducing user fatigue by improving user consistency

Kumara Sastry — Thu, 15 Jun 2006 19:11:05 +0000

Alías, F., Llorà, X., Formiga, L., Sastry, K., Goldberg, D. E. (2006). IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2006). 1, 865—868. [Full paper - PDF].

Abstract:
The quality of corpus-based text-to-speech systems depends on the accuracy of the unit selection process, which in turn relies on the cost function definition. This function should map the user perceptual preference when selecting synthesis units, which is a very difficult task. This paper continues our previous work on fusing the human judgements with the cost function by means of interactive weight tuning. The application of active interactive genetics algorithms mitigates user fatigue by improving user consistency. As a result, the obtained weights generate more natural synthetic speech when compared to previous objective and subjective proposals.

Evaluation consistency in iGAs: User contradictions as cycles in partial-ordering graphs

Kumara Sastry — Wed, 11 Jan 2006 15:05:32 +0000

Llorà, X., Alías, F. , Formiga, L., Sastry, K., Goldberg, D. E. (2005). IlliGAL Report No. 2005022. University of Illinois at Urbana Champaign. [Full paper - PDF] [Full paper - PS].

Abstract:
Active interactive genetic algorithms (aiGAs) rely on actively optimizing synthetic fitness functions. In interactive genetic algorithms (iGAs) framework, user evaluations provide the necessary input for synthesizing a reasonably accurate surrogate fitness function that models user evaluations or, in other words, his/her decision preferences. User evaluations collected via tournament selection only provide partial-ordering relations between solutions. Active iGAs assemble a partial-ordering graph of user evaluations. In such a directed graph, any contradictory evaluation provided by the user introduces a cycle in the graph. This property is explored in this paper to measure the consistency of the evaluations provided by the user along the evolutionary process. The consistency measures are applied to a real-world problem, the weight tuning of the cost function involved in corpus-based text-to-speech synthesis. Results show the usefulness of such measures to identify inconsistent users during the evolutionary tuning process, and successfully the number of evaluations required by more than half.

Combating user fatigue in iGAs: Partial ordering, support vector machines, and synthetic fitness

Kumara Sastry — Fri, 24 Jun 2005 17:00:21 +0000

Llorà, X., Sastry, K., Goldberg, D. E., Gupta, A., Lakshmi, L. (2005). Proceedings of the Genetic and Evolutionary Computation Conference. 1363—1370. [Full paper - PDF] [Full paper - PS] [Presentation slides].

Abstract:
One of the daunting challenges of interactive genetic algorithms (iGAs)—genetic algorithms in which fitness measure of a solution is provided by a human rather than by a fitness function, model, or computation—is user fatigue which leads to sub-optimal solutions. This paper proposes a method to combat user fatigue by augmenting user evaluations with a synthetic fitness function. The proposed method combines partial ordering concepts, notion of non-domination from multiobjective optimization, and support vector machines to synthesize a fitness model based on user evaluation. The proposed method is used in an iGA on a simple test problem and the results demonstrate that the method actively combats user fatigue by requiring 3—7 times less user evaluation when compared to a simple iGA.