Partially Ordered Preferences in Decision Trees: Computing Strategies with Imprecision in Probabilities

Partially ordered preferences generally lead to choices that do not abide by standard expected utility guidelines; often such preferences are revealed by imprecision in probability values. We investigate five criteria for strategy selection in decision trees with imprecision in probabilities: “extensive” Γ-maximin and Γ-maximax, interval dominance, maximality and E-admissibility. We present algorithms that generate strategies for all these criteria; our main contribution is an algorithm for Eadmissibility that runs over admissible strategies rather than over sets of probability distributions.

Privacy-Protected Facial Biometric Verification via Fuzzy Forest Learning

Although visual surveillance has emerged as an effective technolody for public security, privacy has become an issue of great concern in the transmission and distribution of surveillance videos. For example, personal facial images should not be browsed without permission. To cope with this issue, face image scrambling has emerged as a simple solution for privacyrelated applications. Consequently, online facial biometric verification needs to be carried out in the scrambled domain thus bringing a new challenge to face classification. In this paper, we investigate face verification issues in the scrambled domain and propose a novel scheme to handle this challenge. In our proposed method, to make feature extraction from scrambled face images robust, a biased random subspace sampling scheme is applied to construct fuzzy decision trees from randomly selected features, and fuzzy forest decision using fuzzy memberships is then obtained from combining all fuzzy tree decisions. In our experiment, we first estimated the optimal parameters for the construction of the random forest, and then applied the optimized model to the benchmark tests using three publically available face datasets. The experimental results validated that our proposed scheme can robustly cope with the challenging tests in the scrambled domain, and achieved an improved accuracy over all tests, making our method a promising candidate for the emerging privacy-related facial biometric applications.

A decision support system for managing irrigation in agriculture

In this paper, an automatic Smart Irrigation Decision Support System, SIDSS, is proposed to manage irrigation in agriculture. Our system estimates the weekly irrigations needs of a plantation, on the basis of both soil measurements and climatic variables gathered by several autonomous nodes deployed in field. This enables a closed loop control scheme to adapt the decision support system to local perturbations and estimation errors. Two machine learning techniques, PLSR and ANFIS, are proposed as reasoning engine of our SIDSS. Our approach is validated on three commercial plantations of citrus trees located in the South-East of Spain. Performance is tested against decisions taken by a human expert.

Rapid tryptophan depletion improves decision-making cognition in healthy humans without affecting reversal learning or set shifting

Rapid tryptophan (Trp) depletion (RTD) has been reported to cause deterioration in the quality of decision making and impaired reversal learning, while leaving attentional set shifting relatively unimpaired. These findings have been attributed to a more powerful neuromodulatory effect of reduced 5-HT on ventral prefrontal cortex (PFC) than on dorsolateral PFC. In view of the limited number of reports, the aim of this study was to independently replicate these findings using the same test paradigms. Healthy human subjects without a personal or family history of affective disorder were assessed using a computerized decision making/gambling task and the CANTAB ID/ED attentional set-shifting task under Trp-depleted (n=17; nine males and eight females) or control (n=15; seven males and eight females) conditions, in a double-blind, randomized, parallel-group design. There was no significant effect of RTD on set shifting, reversal learning, risk taking, impulsivity, or subjective mood. However, RTD significantly altered decision making such that depleted subjects chose the more likely of two possible outcomes significantly more often than controls. This is in direct contrast to the previous report that subjects chose the more likely outcome significantly less often following RTD. In the terminology of that report, our result may be interpreted as improvement in the quality of decision making following RTD. This contrast between studies highlights the variability in the cognitive effects of RTD between apparently similar groups of healthy subjects, and suggests the need for future RTD studies to control for a range of personality, family history, and genetic factors that may be associated with 5-HT function.

What is a reasonable time from decision to delivery by Caesarean section? Evidence from 533 deliveries.

Objective To determine how long it takes from the decision to achieve delivery by non-elective caesarean section (DDI), the influence on this interval, and the impact on neonatal condition at birth. Design Twelve months prospective data collection on all non-elective caesarean sections. Methods Prospective collection of data relating to all caesarean sections in 1996 in a major teaching hospital obstetric unit was conducted, without the knowledge of the other clinicians providing clinical care. Details of the indication for section, the day and time of the decision and the interval till delivery were recorded as well as the seniority of the surgeon, and condition of the baby at birth. Results The mean time from decision-to-delivery for 100 emergency intrapartum caesarean sections was 42.9 minutes for fetal distress and 71.1 minutes for 230 without fetal distress (P<0.0001). For 22 'crash' sections the mean time from decision-to-delivery was 27.4 minutes; for 13 urgent antepartum deliveries for fetal reasons it was 124.7 minutes and for 21 with maternal reasons it was 97.4 minutes. The seniority of the surgeon managing the patient did not appear to influence the interval, nor did the time of day or day of the week when the delivery occurred. Intrapartum sections were quicker the more advanced the labour, and general anaesthesia was associated with shorter intervals than regional anaesthesia for emergency caesarean section for fetal distress (P<0.001). Babies born within one hour of the decision tended to be more acidaemic than those born later, irrespective of the indication for delivery. Babies tended to be in better condition when a time from decision-to-delivery was not recorded than those for whom the information had been recorded. Conclusion Fewer than 40% intrapartum deliveries by caesarean section for fetal distress were achieved within 30 minutes of the decision, despite that being the unit standard. There was, however, no evidence to indicate that overall an interval up to 120 minutes was detrimental to the neonate unless the delivery was a 'crash' caesarean section. These data thus do not provide evidence to sustain the recommendation of a standard of 30 minutes for intrapartum delivery by caesarean section.