黃冠華

Paired Chest X-Ray Image Translation for Predicting Treatment Response: A Feasibility Study under Real-World Clinical Data Challenges

數據科學與工程研究所

胸部X光影像、深度學習、治療反應預測、條件式生成對抗網路、擴散模型、兩階段串接架構、殘差注意力模組、Chest X-ray Image Translation、Deep Learning、Treatment Response Prediction、Conditional Generative Adversarial Network、Diffusion Model、Two-Stage Cascaded Architecture、Residual Attention Module

On Pretraining Granularity and Pseudo Box Quality for Chest X-ray Abnormality Localization

數據科學與工程研究所

胸部X光影像、自監督學習、物件偵測、異常定位、偽邊界框、醫學影像預訓練、chest X-ray、self-supervised learning、object detection、abnormality localization、pseudo bounding boxes、medical image pretraining

A Domain-Incremental Study of Cross-Dataset Colon Polyp Segmentation

統計學研究所

大腸息肉影像分割、持續學習、低秩適應、專家混合、跨資料集適應、醫學影像分割、colon polyp segmentation、continual learning、low-rank adaptation、mixture of experts、cross-dataset adaptation、medical image segmentation

Modeling Label Dependency, Disease-specific Features, and Latent Structures for Chest X-ray Multi-label Disease Classification

統計學研究所

胸腔X光、多標籤分類、標籤相依性、疾病專屬特徵、潛在結構、動態專家模型、Chest X-ray、multi-label classification、label dependency、disease-specific features、latent structure、dynamic expert modeling

Development and Application of a Breath Volatile Organic Compound Analysis Workflow and Interactive Shiny Platform

統計學研究所

呼出氣體分析、揮發性有機化合物、Shiny平台、背景校正、測量誤差校正、變數選擇、線性混合效應模型、breath analysis、volatile organic compounds、Shiny platform、background correction、measurement error correction、variable selection、linear mixed-effects model

More Efficient Empirical P-values for Machine Learning Approaches

公共衛生研究所・共同指導：郭炤裕

機器學習、排列檢定、序列蒙地卡羅、經驗p值、計算效率、Machine Learning、Permutation Testing、Sequential Monte Carlo、Empirical P-Value、Computational Efficiency

Enhancing Multi-Label Classification via Novel Loss Functions

公共衛生研究所・共同指導：郭炤裕

深度學習、多標籤分類、損失函數、實例標籤資訊、模型效能優化、Deep Learning、Multi-Label Classification、Loss Function、Instance-Level Label Information、Model Performance Optimization

Integrating Latent Class Modeling into Deep Neural Networks for Learning Label Dependencies in Multi-Label Image Classification

統計學研究所

多標籤分類、潛在類別分析、共現矩陣、電腦視覺、醫學影像、Multi-label classification、Latent class analysis、Co-occurrence embedding、Computer vision、Medical image

CNN vs. Transformer: A Comprehensive Comparison and Analysis of Deep Learning Models for Medical Imaging Tasks

數據科學與工程研究所

醫學影像分析、卷積神經網路、視覺變換器、自注意力機制、多標籤分類、語意分割、物件偵測、特徵表徵、對抗穩健性、Medical image analysis、Convolutional neural network、Transformer、Self-attention、Multi-label classification、Semantic segmentation、Object detection、Feature representation、Adversarial robustness

Self-Supervised Learning with Domain-Informed Pseudo Labels for Abnormality Localization in Chest X-rays

數據科學與工程研究所

胸部X光影像、深度學習、物件偵測、自監督學習、偽邊界框生成、chest X-ray image、deep learning、object detection、self-supervised learning、pseudo bounding boxes generation

Empirical and Analytical P-values of the eXtreme Gradient Boosting Machine in Regression Problems

公共衛生研究所・共同指導：郭炤裕

機器學習、模型評估、顯著性、線性迴歸、相關係數、Machine Learning、Model Evaluation、Significance、Linear Regression、Pearson's correlation

Adapting Lightweight Language Models to the Statistical Domain: A Study of Multi-Stage Fine-Tuning

統計學研究所

大型語言模型、領域適應、多階段訓練策略、直接偏好優化、統計語言模 型、參數高效微調、Large Language Models、Domain Adaptation、Multi-Stage Training Strategy、Direct Preference Optimization (DPO)、Statistical Language Models、Parameter-Efficient Fine-Tuning (PEFT)

Deep Learning for Paired Chest X-Ray Image Analysis: Treatment Response Classification, Generation, and Denoising

統計學研究所

胸部X光影像、深度學習、縱向分析、孿生網路、條件式生成對抗網路、殘差注意力模組、Chest X-rays、Deep learning、Longitudinal analysis、Siamese network、Conditional generative adversarial networks、Residual attention module

A Revolutionary Model Fitting for Machine Learning

公共衛生研究所・共同指導：郭炤裕

機器學習、XGBoost、神經網路、監督式學習、非監督式學習、Machine Learning、eXtreme Gradient Boosting Machine、Neural Network、Supervised Learning、Unsupervised Learning

A Novel Structured Masking Algorithm for Variational Auto- Encoder (VAE) with Image-Transformed Medical Data

公共衛生研究所・共同指導：郭炤裕

遺失資料、深度學習、變分⾃動編碼器、遮罩演算法、結構式插補、Missing Data、Deep Learning、Variational Autoencoder (VAE)、Masking Algorithm、Structured Imputation

Novel Evaluation Metrics for Continuous Features in Machine Learning

公共衛生研究所・共同指導：郭炤裕

機器學習、模型評估、回歸預測、多數準確正確率、混淆矩陣、machine learning、model evaluation、regression、major accuracy、confusion matrix

Machine Learning Models Using Volatile Organic Compounds for Disease Prediction: Background Correction and Variable Selection

統計學研究所

呼氣分析、揮發性有機化合物、極限梯度提升、線性混合效應模型、背景校正、變數選擇、Breath Analysis、Volatile Organic Compound、Extreme Gradient Boosting、Linear Mixed Effects Model、Background Correction、Variable Selection

Interference of low-frequency words on self-attention encoding

數據科學與工程研究所

自注意力、詞嵌入模型、資訊檢索、雙向編碼器、點間互資訊、self-attention、word embedding model、information retrieval、bidirectional encoder、point mutual information

An implementation of multiple object detection strategies for disease detection in chest X-ray images

統計學研究所

深度學習、電腦視覺、物件偵測、少樣本物件偵測、醫學影像、deep learning、computer vision、object detection、few-shot object detection、medical image

Self-supervised Pre-training for the Multi-label Classification of Chest X-ray Images

數據科學與工程研究所

自我監督學習、深度學習、胸部X光影像、多標籤分類、預訓練、Self-supervised learning、Deep learning、Chest X-rays、Multi-label classification、Pre-training

IMPROVE: Integrate Multi-Pattern Resolution Of Variant Encoders to enchance model performance

統計學研究所

深度集成、多標籤分類、息肉分割、加權損失函數、醫學影像分析、DeepEnsemble、multi-label classification、polyp segmentation、weighted loss function、medical image analysis

Statistical Inferences for Conditionally Dependent Latent Class Models with Covariates

統計學研究所

潛在類別模型、廣義動差估計、迭代比例擬合算法、latent class models、generalized method of moments、iterative proportional fitting

Automated Convolutional Neural Network and Transformer for Multi-Class Classification of Three-Dimensional Brain Images

數據科學與工程研究所

帕金森氏症、單光子電腦斷層掃描、深度學習、Transformer、自動化機器學習、Parkinson's disease、single photon emission computed tomography、deep learning、automated machine learning

Using endophenotype to detect the gene-gene interaction in drug sensitivity trials

統計學研究所

因果推論、中介效應分析、基因表現量、穩定表現型、藥物敏感性試驗、基因與基因交互作用、causal inference、drug sensitivity、endophenotype、gene expression、mediation analysis、gene-gene interaction

Common thorax disease detection in chest X-rays images using deep learning-based object detection models

統計學研究所

胸部X光影像、深度學習、物件偵測、二元分類、對比度增強、chest X-rays、deep learning、object detection、binary classification、contrast enhancement

Deep transfer learning on multi label classification of chest x ray images

統計學研究所

胸部X光影像、深度學習、遷移學習、多標籤分類、chest x-rays、deep learning、transfer learning、multi label classification

Deep Convolutional Neural Networks for Multi­-Class Classification of Three Dimensional Brain Images

統計學研究所

帕金森氏症、單光子電腦斷層掃描、深度學習、共同訓練、序列分類、Parkinson’s Disease、SPECT、Deep Learning、Co-Train、Sequence Classification

A new summary on the aliasing properties of factorial and supersaturated designs

統計學研究所・共同指導：潘建興

廣義字長模式、廣義解析度、非正規因子設計、超飽和設計、generalized wordlength pattern、generalized resolution、nonregular fractional factorial design、supersaturated design

Segmentation of Low-Grade Gliomas using Deep U-Net with Transfer Learning

統計學研究所・共同指導：NurIriawan

低度惡性神經膠質瘤、腦瘤影像分割、U-Net、VGG16、遷移學習、深度學習、low-grade gliomas、brain tumor segmentation、transfer learning、deep learning

Combining Knowledge Graph and Text for Training Domain-specific Word Embedding

數據科學與工程研究所

自然語言處理、知識圖譜、語料庫、詞向量、特定領域詞向量、Natural Language Processing、Knowledge Graph、Corpus、Word Embedding、Domain-specific word embedding

Bayesian Optimization and Machine Learning Classification for Hyper-parameter Selection in Latent Class Regression Models

統計學研究所

貝氏優化、分類、超參數選擇、潛在類別迴歸模型、機器學習、Bayesian optimization、Classification、hyper-parameter selection、latent class regression model、machine learning

Multi-label deep learning classification of chest x-rays for common thorax disease prediction

統計學研究所

胸部X光片、深度學習、遷移學習、多標籤分類、chest x-rays、deep learning、transfer learning、multi-label classification

ECG Wave and Cardiac Cycle Analysis Based on Deep Learning Method

資訊科學與工程研究所・共同指導：荊宇泰

心電圖、生醫訊號、深度學習、穿戴式裝置、Electrocardiography、u-net、Deep learning、wearable device、Biomedical signal

Multiclass machine learning classification of functional brain images

統計學研究所

帕金森氏症、單光子電腦斷層掃描、主成分分析、多線性主成分分析、圖像統計量、勞斯紋理能量計算、分類模型、整體學習模型、深度學習模型、VGG16、自動化機器學習模型、Parkinson’s disease、Single Photon Emission Computed Tomography (SPECT)、Principal Component Analysis (PCA)、Multilinear principal component analysis (MPCA)、image statistics、Laws' Texture Energy Measure (LTEM)、classification model、ensemble learning、deep learning、Auto-ML

Machine Learning Classification for Hyper-parameter Selection in Latent Class Regression Models

統計學研究所

分類、超參數選擇、潛在類別迴歸模型、機器學習、classification、hyper-parameter selection、latent class regression model、machine learning

Machine learning classification of SPECT functional brain imaging for Parkinson's disease stage prediction

統計學研究所

帕金森氏症、分類模型、整體學習、深度學習、Parkinson's Disease、classification model、ensemble learning、deep learning

Development and Application of Automatic Statistical Analysis System

統計學研究所

自動分析系統、集群分析、探索性數據分析、迴歸分析、Automatic analysis system、clustering analysis、exploratory data analysis、regression analysis

Predicting AMC Analytical Service Payment with Machine Learning

理學院應用科技學程

資料探勘、機械學習、AMC分析、貝氏分類、隨機森林、類神經網路、支持向量機、Data Mining、Machine Learning、AMC analysis、Naive Bayes、Random Forest、Neural Network、Support Vector Machine

TrackNet: Tennis Ball Tracking from Broadcast Video by Deep Learning Networks

資訊科學與工程研究所・共同指導：易志偉

球類追蹤、網球、深度學習、熱度圖、轉播、ball tracking、tennis、deep learning、heatmaps、broadcast

A Modified Generalized Estimating Equation (GEE) Approach for Latent Class Models with Covariate Effects on Measured and Underlying Variables

統計學研究所・共同指導：陳志榮

潛在類別迴歸、廣義季方程式、工作共變量、高斯-牛頓法、regression extension of latent class analysis、generalized estimating equation、working covariance、Gauss-Newton method

A goodness-of-fit test procedure for the latent class regression model

統計學研究所・共同指導：陳志榮

類別資料、潛在類別迴歸模型、廣義概似比率檢定、適合度、categorical data、latent class regression、generalized likelihood ratio test、goodness-of- fit

On Model Diagnostics of Performance Measures in Classification Models and Its Application to Active Learning

統計學研究所・共同指導：張源俊

接受者操作特徵曲線、接受者操作特徵曲線下面積、區域影響、影響函數、不平衡資料、累計提升表、主動學習、表現測度、二元分類、Receiver operating characteristic (ROC) curve、Area under ROC curve (AUC)、Local influence、Influence function、Imbalanced data、Cumulative lift charts、Active learning、Performance measures、Binary classification

Variable selection in the latent class model and its application in traffic flow prediction

統計學研究所

迴歸潛在類別模型、高維度資料、變數選擇、交替K均值分群法、交通流量預測、regression extend of latent class analysis、high-dimensional data、variable selection、alternate k-means clustering、traffic flow prediction

Ensemble learning for remaining useful life prediction of equipment components

統計學研究所

整體學習法、靴拔重抽總合法、推升法、堆疊法、移動區塊靴拔法、剩餘壽命、機台設備零件、感測資料、線性混和效應、參數衰變模型、ensemble learning、bagging、boosting、stacking、moving block bootstrap、remaining useful life、equipment components、censoring data、linear mixed effects、parametric degradation model

The abnormal manufacturing process diagnosis and aging process analysis on MOCVD MFC component data

統計學研究所

MFC、假設檢定、羅吉斯廻歸、hypothesis testing、logistic regression

Linear Mixed Effects Models for Data Analysis and Error Prediction of MOCVD Particle Filter Components

統計學研究所

MOCVD機台、過濾器、變數選擇、線性混合效用模型、預測、F檢定、概似比檢定、AIC、BIC、Conditional AIC、Lasso、交叉分析、MOCVD、Particle Filter、variable selection、linear mixed effect model、prediction、F-test、likelihood ratio test、cross-validation

Comparison of Several RNA-Seq Expression Analytic Tools

統計學研究所・共同指導：劉玉麗

基因表現量的核糖核酸序列分析、選擇性剪切、單純總片斷計數方法、考慮不同轉錄結構下之計數方法、RNA-Seq、alternative splicing、raw-count method、isoform deconvolution method

Approaches and Comparison for Tests of Association with Rare Variants

統計學研究所

罕見變異關聯性分析、次世代定序、遺傳性疾病、Rare Variants Association Study、Next-Generation Sequencing (NGS)、Hereditary Disease

Association Studies of Short DNA Fragments from Next-Generation Sequencing

統計學研究所

次世代基因定序、UK10K、邏輯斯迴歸、卜瓦松迴歸、Next-generation sequencing、logistic regression、Poisson regression

A Multifunctional Bayesian Procedure for Detecting Copy Number Variations from Sequencing Read Depths

統計學研究所

貝氏推論、拷貝數變異、次世代定序、可逆躍式馬可夫鏈蒙地卡羅法、Bayesian inference、Copy number variations、next generation sequencing、reversible jump Markov chain Monte Carlo

Method comparison for discovering copy number variation with next-generation sequencing in target exome regions

統計學研究所

次世代基因定序、外顯子組測拷貝數變異、拷貝數變異、cn.MOPS、CoNIFER、ExomeDepth、exomeCopy、Next-generation sequencing、exome sequencing、copy number variations

Method comparison for discovering copy number variation with next generation sequencing in whole genome

統計學研究所

拷貝數變異、全基因體定序、Copy number variation、whole genome sequencing、cn.MOPS、CNVnator、JointSLM、RDXplorer、ReadDepth

Current approaches for tests of association with rare variants

統計學研究所

全基因組關聯性分析、次世代定序、罕見變異、遺傳性疾病、Genome-Wide Association Study (GWAS)、Next-Generation Sequencing (NGS)、Rare Variants、Hereditary Disease

The Workflow for Next-Generation Sequencing Data Analysis of Human Genome

統計學研究所

全基因體定序、外顯子定序、新一代全基因定序、BWA、Bowtie2、Samtools、Picard、GATK

Genotype Imputation Accuracy with Different Reference Panels

統計學研究所

基因型設算、BEAGLE、IMPUTE v2、Genotype Imputation

Genome-Wide Association for Schizophrenia ─ Patient Subgrouping, Genotype Imputation and Genotype Imputation and Genotype Imputation and Genotype Imputation and Rare Rare-variant Test variant Test variant Test variant Testvariant

統計學研究所

全基因相關性分析、精神分裂症、病患分群、基因型設算、GWAS、Regression Extension of Latent Class Analysis、Schizophrenia、Patient Subgroup、Genotype Imputation、Cumulative Minor-allele Test

Bayesian Inferences of Finite Mixture Models: Approaches for an Unknown Number of Components and Label Switching

統計學研究所

標籤互換現象、貝氏分析、混合分配、可逆躍式馬可夫鏈蒙第卡羅法、馬可夫鏈蒙第卡羅法、潛在類別分析、Label switching、Mixture model、Markov chain Monte Carlo、Latent Class Analysis、Bayesian analysis、Reversible jump、transdimensional、identifiability

A Bayesian clustering approach for detecting gene-gene interactions in high-dimensional genotype data

統計學研究所

Dirichlet process mixtures、交互作用項、排序檢定、隨機搜索、Epistasis、Permutation test、Stochastic search