{"responseHeader":{"status":0,"QTime":6,"params":{"q":"{!q.op=AND}id:\"102818\"","hl":"true","hl.simple.post":"","hl.fragsize":"5000","fq":"!embargo_tdt:[NOW TO *]","hl.fl":"ocr_t","hl.method":"unified","wt":"json","hl.simple.pre":""}},"response":{"numFound":1,"start":0,"docs":[{"file_name_t":"Soller-Automated_Detection.pdf","thumb_s":"/08/5b/085b495931455c3e7d63abd8c093bc922b844873.jpg","oldid_t":"compsci 10931","setname_s":"ir_computersa","restricted_i":0,"format_t":"application/pdf","modified_tdt":"2016-05-26T00:00:00Z","file_s":"/c0/d0/c0d0b42aeecea8379736e6a38dd2af930d031459.pdf","title_t":"Page 61","ocr_t":"46 the model utilized both factors and produced 89.0% sensitivity, 88.8% specificity, and 88.9% accuracy. Given all available factors from DRS and MMSE, it chose two from DRS and the first MMSE factor to produce 91.2% sensitivity, 92.0% specificity, and 91.8% accuracy. Assuming a 0.25 a priori probability of a delirious episode, the specificity improves at the expense of the sensitivity. It produced 89% sensitivity, 94% specificity, and 92.8% accuracy. When 10-fold cross-validation was applied, results were similar (i.e., 90% sensitivity, 94% specificity, and 93% accuracy). 8. 7 Stepwise Logistic Regression Model This research project used the maximum likelihood estimation version of logistic regression to compute the parameters of the logistic model [94] [107]. The logistic model estimates the probability of success, a positive response for delirium, as follows p = exp (u) . (1 + exp (u)) (8.28) This can be rewritten as a ratio describing the odds of the event p (1 _ P) = exp (u). (8.29) Taking the logarithm of both sides, u is expressed as the logarithm of the odds, where p u = bo + 2: biXi. (8.30) i=l The advantage of this formulation for u is its linearity, which eases the calculations. The bi are calculated from the regression. Unlike linear discriminant classifiers, no assumption of normality is necessary. At each step in the stepwise process, a dependent variable can be added or removed based on improvement in estimated values of the x2 statistic.","id":102818,"created_tdt":"2016-05-26T00:00:00Z","parent_i":102961,"_version_":1642982670126284801}]},"highlighting":{"102818":{"ocr_t":[]}}}