1. Title Postoperative Patient Data 2. Sources http://www1.ics.uci.edu/~mlearn/MLSummary.html [postoperative-patient-data] 3. Number of Instances: 90 4. Number of Attributes: 8 5. Relevant Information: The classification task of this database is to determine where patients in a postoperative recovery area should be sent to next. Because hypothermia is a significant concern after surgery (Woolery, L. et. al. 1991), the attributes correspond roughly to body temperature measurements. Results: -- LERS (LEM2): 48% accuracy 6. Attribute Information: Variable Size --------------------------- Postoperative 90x8 Class 90x1 Postoperative: 1. L-CORE (patient's internal temperature in C): 3 - high (> 37), 2 - mid (>= 36 and <= 37), 1 - low (< 36) 2. L-SURF (patient's surface temperature in C): 3 - high (> 36.5), 2 - mid (>= 36.5 and <= 35), 1 - low (< 35) 3. L-O2 (oxygen saturation in %): 4 - excellent (>= 98), 3 - good (>= 90 and < 98), 2 - fair (>= 80 and < 90), 1 - poor (< 80) 4. L-BP (last measurement of blood pressure): 3 - high (> 130/90), 2 - mid (<= 130/90 and >= 90/70), 1 - low (< 90/70) 5. SURF-STBL (stability of patient's surface temperature): 3 - stable, 2 - mod-stable, 1 - unstable 6. CORE-STBL (stability of patient's core temperature) 3 - stable, 2 - mod-stable, 1 - unstable 7. BP-STBL (stability of patient's blood pressure) 3 - stable, 2 - mod-stable, 1 - unstable 8. COMFORT (patient's perceived comfort at discharge, measured as an integer between 0 and 20) Class: 1. decision ADM-DECS (discharge decision): 0 - I (patient sent to Intensive Care Unit), 1 - S (patient prepared to go home), 2 - A (patient sent to general hospital floor) 7. Missing attribute values: Attribute 8 has 3 missing values 8. Class Distribution: Class Value Number of instances 0(I) 2 1(S) 24 2(A) 64