Extreme Logistic Regression: A Large Scale Learning Algorithm with Application to Prostate Cancer Mortality Prediction

摘要: With the recent popularity of electronic medical records, enormous amount data is being generated every day at an exponential rate.Machine learning methods have been shown in many studies to be capable producing automatic diagnostic models such as automated prognostic models. However, powerful machine algorithms support vector (SVM), Random Forest (RF) or Kernel Logistic Regression (KLR) are unbearably slow for very large datasets. This makes their use research limited small medium scale problems.This study motivated by ongoing on prostate cancer mortality prediction a national representative US population where SVM and RF took several hours days trainwhereas simple linear logistic regression discriminant analysis take minutes even seconds.Because, most real-world problems non-linear, this paper presents algorithm enabling recently proposed least squares extreme learn The case men diagnosed with early stage provide fast more accurate result than standard statistical methods.