Expertise
- Details
- Category: Kategorie1
- Published: Friday, 01 January 2016 15:48
- Written by Super User
- Hits: 66271
General Statistical methods
- Parametric Methods (ANOVA, regression etc.)
- Non-Parametrics (rank tests, rank correlation etc.)
- Categorical data analysis (Odds ratio, logistic regression)
- Determining of study size, power analysis, sequential studies
- Bootstrap, permutation tests, Monte-Carlo simulation
- Time-series analysis of biosignals (Fourier, Powerspectra, Transferfunction)
- Disease Modeling (Markov chains, trees etc)
Biomarker methods
- Optimized cutoff search
- Graphical displays for exploration (e.g. hazard estimation depending on biomarker)
- Special analysis designs tailored to uncertain biomarkers
- qRT-PCR scoring
- Image analysis/scoring of tissue (FISH, IHC)
- Biological variation, estimation of RCVs
Time-to-Event methods
- Kaplan-Meier estimation, life-table analysis
- Cox-proportional hazard model – conditional hazard ratios
- Frailty analysis, morbidity scores
- Population Evolution Charts (PEC, cutoff-free association analysis)
- Non-parametric hazard estimation, hazard dependent on covariates
Multivariate methods
- Principal Component Analysis, Factor analysis, factor rotation
- Cluster analysis, canonical correlation
- Pattern recognition (Discriminant Analysis, Support Vector Machines etc.)
- Exploratory Analysis e.g. CART
- Trilinear Decomposition (PARAFAC, CANDECOMP, CP)
Home grown approaches (unpublished)
- New time-to-event testing facilities (context PEC)
- Special approaches for estimation of hazard dependencies
- Test Tuning by Optimal Rank ordering (context rank tests)
- Special pre-processing in multivariate separation problems (context pattern recognition, discriminant analysis)
- New algorithmic variants for trilinear modelling
- Tricks in hierarchical cluster analysis