Topic:
Constrained Monotonic Neural Networks


Keywords:
interpretable deep learning,
Model generalization
Neural network,
Monotonicity,

Reference:
Runje, Davor, and Sharath M. Shankaranar…

• Date

  2023-03-17

• Speaker

  김현호

Topic:
RIFS: a randomly restarted incremental feature selection algorithm

keywords:
RIFS
RIFS2D
Feature selection

Reference:
Ye, Y., Zhang, R., Zheng, W. et al. RIFS: a randomly restarted i…

• Date

  2023.02.15

• Speaker

  이형권

Topic:
Image Super-Resolution Using Deep Convolutional Networks

Keywords:
Super-resolution
Deep convolutional neural networks
Sparse coding

Reference:
C. Dong, C. C. Loy, K. He and X. Tang,…

• Date

  2023.02.15

• Speaker

  김주헌

Topic:
Similarity of Neural Network Representations Revisited


Keywords:
Neural Network
Representation Similarity


Reference:
1. Kornblith, Simon, et al. "Similarity of neural network rep…

• Date

  2023-02-01

• Speaker

  김현호

Learning physics constrained dynamic using autoencoder

NeurIPS 2022

• Date

  2023.02.08

• Speaker

  한민석

Paper review
1. Distilling the Knowledge in a Neural Network (G Hinton et al, 2014 NIPS)
2. Self-training with Noisy Student improves ImageNet classification (Q Xie et al, 2020 CVPR)
3. Knowledge …

• Date

  2023.01.18

• Speaker

  김태욱

Topic:
Clustering using silhouette coefficients

Keywords:
Cluster analysis
Cluster validity index
Silhouette coefficient
Objective function clustering

Reference:
F. Batool, "Clustering wit…

• Date

  2023.01.18

• Speaker

  김주헌

Physics-Informed neural network
- Concept & Practice

• Date

  2023.01.10

• Speaker

  한민석

Topic:
Merging Components in Linear Gaussian Cluster-Weighted Models


Keywords:
Mixture Models,
Cluster Weighted Models,
Model based clustering

Reference:
1. Oh, Sangkon, and Byungtae Seo.…

• Date

  2023-01-03

• Speaker

  김현호

Out-of-distribution detection using union of 1-dimensional subspaces

• Date

  2022.11.24

• Speaker

  김태욱

Topic:
LIME & SHAP


Keywords:
Explanable AI
Model-agnostic local surrogate model
Shapley value,
LIME,

Reference:
1. Ribeiro, Marco Tulio, Sameer Singh, and Carlos Guestrin. "" Why should …

• Date

  2022-11-17

• Speaker

  김현호

Topic:
Incorporating Prior Domain Knowledge into Loss Function

Keywords:
Noisy Data
Domain Knowledge
Neural Networks
Deep Learning
Limited Training Data

Reference:
N. Muralidhar, M. R. Is…

• Date

  2022.11.03

• Speaker

  김주헌