Filetype
Neural networks learn knowledge from data. For a monolithic structure, this knowledge can be easily used but not isolated. The many degrees of freedom while learning make knowledge extraction a computationally intensive process as the representation is not unique. Where existing knowledge is inserted to initialize the network for training, the effect becomes subsequently randomized within the solu
The traditional sheet metal forming process, including cast steel die tool and conventional work materials, is fairly well established and analysed. Developments of new alternative materials make great demands upon new forming processes and manufacturing methods. A general goal for the manufacturing industries is to reduce the weight of metal parts, reduce batch sizes and reduce the lead-time of n
This dissertation, based on six included papers, theoretically investigates properties of quantum few-particle systems. An overview of related experimental research - ultra-cold trapped dilute gases, and electrons in quantum dots - is given, followed by a description of some of the studied many-particle phenomena - Bose-Einstein condensation, quantized vortices, Wigner localization and the Tonks-G
The effect of airborne particulate matter on human health is today of considerable research interest. Essential parameters include elemental composition sampling with low pressure cascade impactors and analysis by particle induced x-ray emission spectroscopy (PIXE). It is shown that the aerosol generated by arc welding operations is several orders of magnitude denser than aerosols in ambient air.
