生信机器学习平台

目前使用多种机器学习算法寻优，是生信分析高分文章的常用思路。如最近发表在《Nature Communications》（IF: 14.919）的文章《Machine learning-based integration develops an immune-derived lncRNA signature for improving outcomes in colorectal cancer》,就使用了101 种机器学习预测模型进行分析，最终发现最佳模型是Lasso 和逐步 Cox回归的组合，该组合模型在所有验证数据集中都具有很高的C-index.

螺旋矩阵公司开发的生信多模型机器学习系统 BioMatrix ，一次性运行上百个模型，并自动进行贝叶斯超参数优化。搜索最优解决方案。以下为部分支持算法，欢迎各位老师，联系使用

• Linear Models
  • Ordinary Least Squares
  • Ridge regression and classification
  • Lasso
  • Multi-task Lasso
  • Elastic-Net
  • Multi-task Elastic-Net
  • Least Angle Regression
  • LARS Lasso
  • Orthogonal Matching Pursuit (OMP)
  • Bayesian Regression
  • Logistic regression
  • Generalized Linear Regression
  • Stochastic Gradient Descent – SGD
  • Perceptron
  • Passive Aggressive Algorithms
  • Robustness regression: outliers and modeling errors
  • Quantile Regression
  • Polynomial regression: extending linear models with basis functions
• Linear and Quadratic Discriminant Analysis
  • Dimensionality reduction using Linear Discriminant Analysis
  • Mathematical formulation of the LDA and QDA classifiers
  • Mathematical formulation of LDA dimensionality reduction
  • Shrinkage and Covariance Estimator
  • Estimation algorithms
• Kernel ridge regression
• Support Vector Machines
  • Classification
  • Regression
  • Density estimation, novelty detection
  • Complexity
  • Tips on Practical Use
  • Kernel functions
  • Mathematical formulation
  • Implementation details
• Stochastic Gradient Descent
  • Classification
  • Regression
  • Online One-Class SVM
  • Stochastic Gradient Descent for sparse data
  • Complexity
  • Stopping criterion
  • Tips on Practical Use
  • Mathematical formulation
  • Implementation details
• Nearest Neighbors
  • Unsupervised Nearest Neighbors
  • Nearest Neighbors Classification
  • Nearest Neighbors Regression
  • Nearest Neighbor Algorithms
  • Nearest Centroid Classifier
  • Nearest Neighbors Transformer
  • Neighborhood Components Analysis
• Gaussian Processes
  • Gaussian Process Regression (GPR)
  • GPR examples
  • Gaussian Process Classification (GPC)
  • GPC examples
  • Kernels for Gaussian Processes
• Cross decomposition
  • PLSCanonical
  • PLSSVD
  • PLSRegression
  • Canonical Correlation Analysis
• Naive Bayes
  • Gaussian Naive Bayes
  • Multinomial Naive Bayes
  • Complement Naive Bayes
  • Bernoulli Naive Bayes
  • Categorical Naive Bayes
  • Out-of-core naive Bayes model fitting
• Decision Trees
  • Classification
  • Regression
  • Multi-output problems
  • Complexity
  • Tips on practical use
  • Tree algorithms: ID3, C4.5, C5.0 and CART
  • Mathematical formulation
  • Minimal Cost-Complexity Pruning
• Ensemble methods
  • Bagging meta-estimator
  • Forests of randomized trees
  • AdaBoost
  • Gradient Tree Boosting
  • Histogram-Based Gradient Boosting
  • Voting Classifier
  • Voting Regressor
  • Stacked generalization
• Multiclass and multioutput algorithms
  • Multiclass classification
  • Multilabel classification
  • Multiclass-multioutput classification
  • Multioutput regression
• Feature selection
• Semi-supervised learning
  • Self Training
  • Label Propagation
• Isotonic regression
• Probability calibration
  • 1.16.1. Calibration curves
  • 1.16.2. Calibrating a classifier
• 1.17. Neural network models
  • Multi-layer Perceptron
  • Classification
  • Regression
  • Regularization