Sökning: "Speed gradient"

Visar resultat 1 - 5 av 64 uppsatser innehållade orden Speed gradient.

1. 1. A Gradient Boosting Tree Approach for Behavioural Credit Scoring

   Master-uppsats, KTH/Matematisk statistik

   Författare :Axel Dernsjö; Ebba Blom; [2023]
   Nyckelord :Machine learning; Random forest; Uncertainty measure; Material development; Empirical Bayes; Maskininlärning; Random forest; Osäkerhetsmått; Materialutveckling; Empirical Bayes;

   Sammanfattning : This report evaluates the possibility of using sequential learning in a material development setting to help predict material properties and speed up the development of new materials. To do this a Random forest model was built incorporating carefully calibrated prediction uncertainty estimates. LÄS MER

3. 2. Estimation of Voltage Drop in Power Circuits using Machine Learning Algorithms : Investigating potential applications of machine learning methods in power circuits design

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

   Författare :Dimitrios Koutlis; [2023]
   Nyckelord :Voltage drop estimation; Application-specific Integrated Circuits ASICs ; Machine learning algorithms; XGBoost; Convolutional Neural Networks; Graph Neural Networks; Power circuit optimization; Uppskattning av spänningsfall; applikationsspecifika integrerade kretsar ASIC ; maskininlärningsalgoritmer; XGBoost; konvolutionella neurala nätverk; optimering av strömkretsar;

   Sammanfattning : Accurate estimation of voltage drop (IR drop), in Application-Specific Integrated Circuits (ASICs) is a critical challenge, which impacts their performance and power consumption. As technology advances and die sizes shrink, predicting IR drop fast and accurate becomes increasingly challenging. LÄS MER

4. 3. GONet: Gradient Oriented Fuzzing for Stateful Network Protocol : Improving and Evaluating Fuzzing Efficiency of Stateful Protocol by Mutating Based on Gradient Information

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

   Författare :Quanyu Tao; [2023]
   Nyckelord :Fuzzing; Stateful Protocol; Gradient Oriented; Neural Network;

   Sammanfattning : Network protocol plays a crucial role in supporting a wide range of critical services, of which robustness and reliability are vital. Fuzzing, or fuzz testing, serves as an effective technique to uncover vulnerabilities in software programs. However, fuzzing becomes more complicated when dealing with network protocols due to their massive state. LÄS MER

5. 4. Shoppin’ in the Rain : An Evaluation of the Usefulness of Weather-Based Features for an ML Ranking Model in the Setting of Children’s Clothing Online Retailing

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

   Författare :Isac Lorentz; [2023]
   Nyckelord :Statistical analysis; regression analysis; recommender systems; ensemble learning; electronic commerce; LightGBM; learning to rank; feature selection; weather-based features; fashion; Statistisk analys; regressionsanalys; rekommendationssystem; ensemble-inlärning; näthandel; LightGBM; learning to rank; variabelselektion; väderbaserade variabler; mode;

   Sammanfattning : Online shopping offers numerous benefits, but large product catalogs make it difficult for shoppers to understand the existence and characteristics of every item for sale. To simplify the decision-making process, online retailers use ranking models to recommend products relevant to each individual user. LÄS MER

7. 5. Reinforcement Learning for the Optimization of Explicit Runge-Kutta Method Parameters

   Kandidat-uppsats, Lunds universitet/Matematik LTH; Lunds universitet/Matematik (naturvetenskapliga fakulteten); Lunds universitet/Matematikcentrum

   Författare :Mélanie Fournier; [2023]
   Nyckelord :reinforcement learning; numerical analysis; Runge-Kutta; policy gradient; REINFORCE; Mathematics and Statistics;

   Sammanfattning : Reinforcement learning is one of the three main paradigms in machine learning, which is increasingly used as a method to approach scientific problems. In this thesis, we introduce and use reinforcement learning to find the optimal parameters of a numerical solver. LÄS MER