Sökning: "Maximum Likelihood"
Visar resultat 36 - 40 av 289 uppsatser innehållade orden Maximum Likelihood.
36. CP Violation Studies in Cascade Decay Sequence
Kandidat-uppsats, Uppsala universitet/KärnfysikSammanfattning : This thesis studies CP violation in the decay of the Ξ− hyperon, also known as the cascade baryon, which decays in the sequence Ξ− → Λπ− → pπ−π−. A difference between the angular distribution of this decay sequence and the angular distribution of the charge conjugate decay sequence ͞Ξ+ →͞Λπ+ → ͞pπ+π+, after taking into account the inversion of the momenta due to the parity operation, is indicative of CP violation. LÄS MER
37. Regularized Fine-tuning Strategies for Neural Language Models : Application of entropy regularization on GPT-2
Master-uppsats, Uppsala universitet/Institutionen för lingvistik och filologiSammanfattning : Deep neural language models like GPT-2 is undoubtedly strong at text generation, but often requires special decoding strategies to prevent producing degenerate output - namely repetition. The use of maximum likelihood training objective results in a peaked probability distribution, leading to the over-confidence of neural networks. LÄS MER
38. Demography of Birch Populations across Scandinavia
Master-uppsats, Uppsala universitet/Växtekologi och evolution; Uppsala universitet/Institutionen för biologisk grundutbildningSammanfattning : Boreal forests are particularly vulnerable to climate change, experiencing a much more drastic increase in temperatures and having a limited amount of more northern refugia. The trees making up these vast and important ecosystems already had to adapt previously to environmental pressures brought about by the repeated glaciations during past ice ages. LÄS MER
39. The Impact of Noise on Generative and Discriminative Image Classifiers
Kandidat-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : This report analyzes the difference between discriminative and generative image classifiers when tested on noise. The generative classifier was a maximum-likelihood based classifier using a normalizing flow as the generative model. In this work, a coupling flow such as RealNVP was used. LÄS MER
40. Deep Learning based Approximate Message Passing for MIMO Detection in 5G : Low complexity deep learning algorithms for solving MIMO Detection in real world scenarios
Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : The Fifth Generation (5G) mobile communication system is the latest technology in wireless communications. This technique brings several advantages, in particular by using multiple receiver antennas that serve multiple transmitters. LÄS MER