Refine
Document Type
- Conference Proceeding (9)
- Article (3)
- Part of a Book (2)
Is part of the Bibliography
- yes (14)
Keywords
- Abwasser-Gas-WP/KM (1)
- Artificial intelligence (1)
- Cement production (1)
- Dynamic processes (1)
- Engineering Conference (1)
- Engineering Conferences (1)
- Machine learning (1)
- Poster (1)
- Robustness (1)
- Soft sensors (1)
Department/institution
We present an approach to feature extraction that is a generalization of the classical linear discriminant analysis (LDA) on the basis of deep neural networks (DNNs). As for LDA, discriminative features generated from independent Gaussian class conditionals are assumed. This modeling has the advantages that the intrinsic dimensionality of the feature space is bounded by the number of classes and that the optimal discriminant function is linear. Unfortunately, linear transformations are insufficient to extract optimal discriminative features from arbitrarily distributed raw measurements. The generalized discriminant analysis (GerDA) proposed in this paper uses nonlinear transformations that are learnt by DNNs in a semisupervised fashion. We show that the feature extraction based on our approach displays excellent performance on real-world recognition and detection tasks, such as handwritten digit recognition and face detection. In a series of experiments, we evaluate GerDA features with respect to dimensionality reduction, visualization, classification, and detection. Moreover, we show that GerDA DNNs can preprocess truly high-dimensional input data to low-dimensional representations that facilitate accurate predictions even if simple linear predictors or measures of similarity are used.
This article is on a university course called Engineering Conferences. It has been developed by the authors and installed as a mandatory part of the curriculum in three master’s degree programs for engineering students. The participants of the course are postgraduates with different nationalities, mostly German, and different technical backgrounds. They study Mechanical Engineering, Simulation and Experimental Technology, or International Business Engineering. The basic idea of the course concept goes far beyond simply teaching the standards of academic writing and skills for using scientific publications. By using a learnercentered approach, the students get engaged in typical activities around an active attendance of a real conference. Students learn to locate the field of their bachelor thesis or project report in the world of research communities, scientific journals and engineering conferences. They learn about conferences matching their bachelor/project topic. They write a paper complying with common academic standards, submit it to a mock-up conference, and review submissions of their fellow students. Students also produce a poster and have to defend it in a poster session held publicly on the campus. Engineering Conferences is a course on scientific communication and presentation that also aims at the development of other skills and competences needed in the world of modern engineering.