Topic: Clustering using silhouette coefficients Keywords: Cluster analysis Cluster validity index Silhouette coefficient Objective function clustering Reference: F. Batool, "Clustering wit…
2023.01.18
김주헌
Topic: Feedback loop in machine learning Keywords: Feedback loop Recommender system Control system Reference: A. Khritankov, "Hidden Feedback Loops in Machine Learning Systems: A Simulation…
2023.01.03
Physics-Informed neural network - Concept & Practice
2023.01.10
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.…
2023-01-03
김현호
Out-of-distribution detection using union of 1-dimensional subspaces
2022.11.24
김태욱
NET-DNF Tabular 데이터에 알맞는 Inductive bias로써, DNNF block을 제안함
2022.11.17
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 …
2022-11-17
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…
2022.11.03
Bayesian Linear Regression and Gaussian process regression - full derivation of Bayesian LR - GPR tutorial
2022.10.06
On Mixup Training: Improved Calibration and Predictive Uncertainty for deep Neural Networks Keyword : Calibration, Uncertainty, mix-up
Topic: Certified Monotonic Neural Networks Keywords: interpretable deep learning, Model generalization Neural network, Monotonicity, Reference: Liu, Xingchao, et al. "Certified monotonic…
2022-09-29
Keywords: pruning, initialization, subspace 적은 수의 파라미터를 가진 pruned model이 덜 학습하여도 original network 보다 좋은 performance를 보여줄 수 있다.
2022-06-23
이주형
Topic: How Domain Adaptation Works Keywords: Domain adaptation Transfer learning Learning theory Sample-selection bias Reference: S. Ben-David, J. Blitzer, K. Crammer, A. Kulesza, F. Perei…
2022.06.23
Importance Estimation for Neural Network Pruning
2022.06.09
김태흥
1. Discriminative feature space embedding using angular margin loss 2. Separate Maximum softmax probability score between ID and OOD using weight decay regularization keywords : Out-of-distributio…
05/26
