Sökning: "Neural dekomposition."

Hittade 4 uppsatser innehållade orden Neural dekomposition..

  1. 1. Spectral Portfolio Optimisation with LSTM Stock Price Prediction

    Master-uppsats, KTH/Matematisk statistik

    Författare :Nancy Wang; [2020]
    Nyckelord :Artificial Neural Network; LSTM; Spectral factor model; Portfolio optimisation; Stock price prediction; Time series analysis; Risk estimation; Spectral risk; Frequency-specific beta decomposition; Artificiella neurala nätverk; LSTM; Spektralfaktormodell; Portföljoptimering; Aktieprispredikering; Tidsserieranalys; Riskestimering; Spektra risk; Frekvensspecifik beta dekomposition;

    Sammanfattning : Nobel Prize-winning modern portfolio theory (MPT) has been considered to be one of the most important and influential economic theories within finance and investment management. MPT assumes investors to be riskaverse and uses the variance of asset returns as a proxy of risk to maximise the performance of a portfolio. LÄS MER

  3. 2. An evaluation of deep neural network approaches for traffic speed prediction

    Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)

    Författare :Cosar Ghandeharioon; [2018]
    Nyckelord :Deep Learning; Regression; Time Series; LSTM; Neural decomposition.; Djupinlärning; Regression; Tidsserier; LSTM; Neural dekomposition.;

    Sammanfattning : The transportation industry has a significant effect on the sustainability and development of a society. Learning traffic patterns, and predicting the traffic parameters such as flow or speed for a specific spatiotemporal point is beneficial for transportation systems. LÄS MER

  4. 3. Establishing Effective Techniques for Increasing Deep Neural Networks Inference Speed

    Master-uppsats, KTH/Skolan för datavetenskap och kommunikation (CSC)

    Författare :Albin Sunesson; [2017]
    Nyckelord :Deep Neural Networks; Machine Learning; Inference Speed; Convolution; Decomposition; Network Branches; Julia; MXNET; GPU; CPU.;

    Sammanfattning : Recent trend in deep learning research is to build ever more deep networks (i.e. increase the number of layers) to solve real world classification/optimization problems. This introduces challenges for applications with a latency dependence. LÄS MER

  5. 4. Mammography Classification and Nodule Detection using Deep Neural Networks

    Master-uppsats, KTH/Numerisk analys, NA

    Författare :Fabian Sinzinger; [2017]
    Nyckelord :;

    Sammanfattning : 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