Sökning: "Kostnadsfunktion"
Visar resultat 16 - 20 av 27 uppsatser innehållade ordet Kostnadsfunktion.
16. Mammography Classification and Nodule Detection using Deep Neural Networks
Master-uppsats, KTH/Numerisk analys, NASammanfattning : Mammographic screenings are the most common modality for an early detection of breast cancer, but a robust annotation of the depicted breast tissue presents an ongoing challenge, even for well-experienced radiologists. Computer-aided diagnosis systems can support the human classification. LÄS MER
17. Trajectory Generation under Metric Interval Temporal Logic Specifications
Master-uppsats, KTH/ReglerteknikSammanfattning : Metric interval temporal logic (MITL) provides a tool to formulate high level tasks in an easyway. This allows to specify behaviour of a dynamical system in a given environment which liesbeyond the scope of simple stabilization. Finding an input of the system which satises the MITLformula, demands new techniques and algorithms. LÄS MER
18. Semi-Supervised Learning with Sparse Autoencoders in Automatic Speech Recognition
Master-uppsats, KTH/Skolan för datavetenskap och kommunikation (CSC)Sammanfattning : This work is aimed at exploring semi-supervised learning techniques to improve the performance of Automatic Speech Recognition systems. Semi-supervised learning takes advantage of unlabeled data in order to improve the quality of the representations extracted from the data. LÄS MER
19. Obstacle avoidance for platforms in three-dimensional environments
Master-uppsats, KTH/Skolan för datavetenskap och kommunikation (CSC)Sammanfattning : The field of obstacle avoidance is a well-researched area. Despite this, research on obstacle avoidance in three dimensions is surprisingly sparse. For platforms which are able to navigate three-dimensional space, such as multirotor UAVs, such methods will become more common. LÄS MER
20. Simulating MPC Controlled Lime Injection for the Flue-gas Treatment at Fortum's Thermal Power Plant
Uppsats för yrkesexamina på avancerad nivå, KTH/ReglerteknikSammanfattning : The aim of this thesis is to investigate the possibility to improve the separation of HCl and SO2 in the dry ue-gas treatment for boiler #3 at Fortum's thermal power plant in Hogdalen; by using a model predictive control instead of a PID controller to govern the slaked lime injection. To achieve this an ARMAX model was derived using MATLAB's System Identi cation Toolbox and measurements of the incoming and outgoing levels of HCl, SO2 and the speed of the injection motor. LÄS MER