6 Hits in 5.2 sec

Multicategory large margin classification methods: Hinge losses vs. coherence functions

Zhihua Zhang, Cheng Chen, Guang Dai, Wu-Jun Li, Dit-Yan Yeung
2014 Artificial Intelligence  
Finally, we develop multicategory large margin classification methods by using a so-called multiclass C-loss.  ...  Corresponding to the three hinge losses, we propose three multicategory majorization losses based on a coherence function.  ...  [30] proposed a smooth loss function that called coherence function for developing binary large margin classification methods.  ... 
doi:10.1016/j.artint.2014.06.002 fatcat:6psadlz6qjbi3koxkfhmphnqye

Reinforcement Learning in Modern Biostatistics: Constructing Optimal Adaptive Interventions [article]

Nina Deliu, Joseph Jay Williams, Bibhas Chakraborty
2022 arXiv   pre-print
We provide the first unified instructive survey on RL methods for building AIs, encompassing both dynamic treatment regimes (DTRs) and just-in-time adaptive interventions in mobile health (mHealth).  ...  Authors propose a robust OWL (ROWL), based on an angle-based classification structure, designed for multicategory classification problems, and a new family of robust loss functions to build more stable  ...  Authors use sequential binary methods by proposing a margin-based learning (build upon the large-margin unified machine (LUM) loss), which has a special case the standard OWL.  ... 
arXiv:2203.02605v1 fatcat:a5m6fa7ec5bznoghkgi3ojo3de

Data Recovery Investigations at the Tank Destroyer Site (41CV1378) at Fort Hood, Coryell County, Texas

Douglas Boyd, John Dockall, Karl Kibler, Gemma Mehalchick, Laura Short
2014 Index of Texas Archaeology Open Access Grey Literature from the Lone Star State  
[Hinge / step] HST 05. [Overshot (outrepasse)] OVR 06. [Material flaw] MFL 07. [Platform loss] PLL 08. [Excessive heating] HTF 09.  ...  Hinge / Step 05. Overshot (outrepasse) 06. Material flaw 07. Platform loss 08. Excessive heating 09. Exhausted 16. Alteration 00. None observed 01. Thermal 02. White patina 03.  ... 
doi:10.21112/ita.2014.1.2 fatcat:i556bnvd7be7vh75z6tzd6p7pm

Learning for the internet : kernel embeddings and optimisation [article]

Novi Quadrianto, University, The Australian National, University, The Australian National
In this thesis, we develop principled machine learning methods suited for complex real-world Internet challenges.  ...  The second part addresses refinements of existing machine learning models and algorithms to scale to large data. The contributions of this thesis include a streaming alg [...]  ...  Since we anticipate the relevant length scale in the margin distribution to be in the order of 1 (after all, we use a loss function, i.e. a hinge loss, which uses a margin of 1) we pick a Gaussian RBF  ... 
doi:10.25911/5d5001fc96254 fatcat:2hteh4qcsnftvbhozwqk443sqq

Discriminating Between Closely Related Languages on Twitter

Anton Železnikar, Nikola Ljubeši´cljubeši´c, Denis Kranjči´kranjči´c
2015 unpublished
We show that by using a simple bag-of-words model, univariate feature selection, 320 strongest features and a standard classifier, we reach user classification accuracy of ∼98%.  ...  SVM (based on liblinear) Discriminant Super-plane separation Minimizing the loss of regular hinge, soft margin maximization Hinge loss Sequential dual method Maximum weighted test sample  ...  separation, kernel trick Minimizing the loss of regular hinge, soft margin maximization Hinge loss Sequential minimal optimization algorithm (SMO) Maximum class of test samples Linear  ... 

Robotic isotropy and optimal robot design of planar manipulators

M.V. Kircanski
Proceedings of the 1994 IEEE International Conference on Robotics and Automation  
Wireless LAN: Infrared vs.  ...  and phase margin via the Nyquist diagram Stability Gain margin and phase margin via the Bode-plots Relation between closed loop transient and closed loop frequency response Relation between closed loop  ...  Frequency-Response Plots Bode plots, Polar plots, Log-magnitude Vs phase plots, Nyquist stability criterion, stability analysis, Relative stability, gain margin, phase margin, stability analysis of system  ... 
doi:10.1109/robot.1994.351213 dblp:conf/icra/Kircanski94 fatcat:uawqnzropbgivcanzqihtimt2m