2016-05-19から1日間の記事一覧

時系列・空間とkrigingとか Time-series/Space and krigign etc.

Conventional analysis depends on independency among values Time series/ spacial data have values that are next to each other are tightly associated and this feature should be absolutely considered. At a glance

MCMC

What is new? This approach requires computers, that was not available in 20th century; that is why MCMC is being used for tasks that could not be solved by the methods before and also MCMC is being used for tasks that have been answered by…

検定からベイズ推定へ Frequentist approach ~ tests and Bayesian approach

日本語説明 reading material on Bayes Statement on p-value (ASA)

Noisy High-Throughput Biological Data

Noisy because Biological phenomena are "noisy" ~ heterogeneity is the important feature of biology. Experiments have many factors that add noise to data. Highthroughput systems realize "highthoughput" by sacrificing preciseness somehow. No…

small n large p

100 samples x 25000 genes You can predict perfectly when you are allowed to use explanatory variables as many as sample size. n <- 10 m <- 10 p <- sample(0:1,n,replace=TRUE) g <- matrix(sample(0:1,n*m,replace=TRUE),n,m) p g lm.out <- lm(p~…

High Dimensionality

Many genes Many features Many biomarkers 検索結果 Sparse cells with 5 markers grid; one cubicle only has 1 cell in average. It is too sparse to estimate density in a regular way. No center, no common individuals N <- 10^4 x <- runif(N)-0.5…

Multiple-Comparison Issue

When you test multiple times, you should not believe nominal p-values of individual tests. 検定を複数行った場合には、個々の検定のp値をそのまま使って解釈できない p <- runif(10^5) hist(p) plot(sort(p),pch=20,cex=0.1) alpha <- 0.05 abline(v=l…

手法・目的の分類オーバービュー〜90分で学ぶ医学・ライフサイエンスの統計学手法のいまどき

Statistical Bioinformatics: For Biomedical and Life Science Researchers(の目次) Quality Control of High-Throughput Data Statistical Tests, Statistical Significance, Error Controlling Classification/Clustering Unsupervised Learning Supervis…

High/multi-dimensional data analysis

Visualization dimension reduction manifold dimension reduction

どんなことが課題か〜90分で学ぶ医学・ライフサイエンスの統計学手法のいまどき

Statistical Bioinformatics: For Biomedical and Life Science Researchers(の目次) Challenge 1 Multiple-Comparison Issue Challenge 2 High-Dimensional Biological Data Challenge 3 Small-n and Large-p Problem Challenge 4 Noisy High-Throughput Bi…

90分で学ぶ医学・ライフサイエンスの統計学手法のいまどき

対象(医科学修士院生1回生) 資料 Statistical Bioinformatics: For Biomedical and Life Science Researchers(の目次) Beyond the hype: Big data concepts, methods, and analytics Big Data Analysis Using Modern Statistical and Machine Learning Meth…