Sensory cut-off point obtained from survival analysis statistics

In the present work we applied interval-censored survival analysis techniques to estimate sensory cut-off points based on consumer’s decision to accept or reject food products taking into account the inherent variability in sensory measurements. We compared the values obtained using this survival analysis methodology with those obtained by applying a previous regression based method. Cut-off point (COP) estimations were made for acid flavor in yogurt, strawberry flavor in a strawberry flavored drink and appearance quality index in broccoli. For two of these products the regression based cut-off points were unrealistic, and would lead to much too conservative COP’s, leading to unnecessary rejection of samples in quality control inspections or very short shelf-lives. For one of the products (strawberry flavored drink), the survival and regression-based COP’s were comparable. The survival analysis methodology is recommended for estimating sensory cut-off points in food products.     

Salmonella enterica serovar Choleraesuis on fresh-cut cucumber slices after reduction treatments

Cucumber is a popular fruit around the world and has been implicated in Salmonella food poisonings. S. Choleraesuis is a serovar that can cause pig and human infections but was rarely examined in food safety context. To investigate S. Choleraesuis behavior on cucumber slices, it was inoculated, at 10⁴ colony forming units (CFU)/mL, onto fresh-cut cucumber slices and subjected to reduction with either high hydrostatic pressurization (HHP), hydrogen peroxide (H₂O₂), or Peredibacter sp. BD2GS treatment, its reduction and survival during 48 h storage at 4 °C and 25 °C were compared.Reduction tests revealed that 5% H₂O₂ was most effective in killing S. Choleraesuis, with 97.5% reduction after 15 min action, compared to 90.7%, 87.7%, 29.2% and 60.2% reduction rates with HHP, 2.5% H₂O₂, high- and low-dose BD2GS treatments, respectively.At the end of storage, contrast to no changes at 4 °C, S. Choleraesuis counts rose significantly (p < 0.05) at 25 °C. Compared to control that reached 7.1 ± 0.1 log CFU/g, HHP, 5% and 2.5% H₂O₂ attained 6.3 ± 0.1, 6.7 ± 0.1 and 6.4 ± 0.2 log CFU/g correspondingly, whereas high- and low-dose BD2GS attained 4.9 ± 0.1 and 5.9 ± 0.1 log CFU/g respectively. A shared growth peak of between 9 h and 12 h was noted in all treatments except high-dose BD2GS where it occurred in the first 3 h. Results of this study revealed the effectiveness of 5% H₂O₂ in the reduction of S. Choleraesuis, and demonstrated that if not stored properly, contaminated cucumber slices, though treated, can still have potentials to cause S. Choleraesuis outbreaks.     

Effect of dehydrated storage on the survival of Francisella tularensis in infant formula

Francisella tularensis is a Gram-negative bacterium that can cause gastrointestinal or oropharyngeal tularemia in humans from ingestion of contaminated food or water. Despite the potential for accidental or intentional contamination of foods with F. tularensis, there are few studies on the long-term survivability of this organism in food matrices. Infant formula has previously been implicated as a vehicle for the transmission of a variety of bacterial pathogens in infants. In this study, we investigated the survival of F. tularensis in dehydrated infant formula under various storage conditions. F. tularensis was stored for up to 12 weeks in dehydrated infant formula in an ambient air, dry or nitrogen atmosphere. Viable counts of fresh F. tularensis at 12 weeks in infant formula revealed a 4.15, 3.37 and 3.72-log decrease in ambient air, dry and nitrogen atmosphere, respectively. D-values were calculated (in weeks) as 3.99, 4.68 and 4.47 in air, dry and nitrogen atmosphere, respectively.     

Effects of storage temperature and water activity on the survival of bifidobacteria in powder form

Effects of water activity and storage temperature on survival of bifidobacteria in powder form were investigated and kinetic analyses were performed to reveal characteristics of the stability. A significant positive correlation was observed between water activity and natural logarithm of the inactivation rate constant of bifidobacteria powder, indicating that higher water activity induced lower stability of bifidobacteria in powder form. Also, higher temperature condition induced lower survival rate, which was supported by that the stability was followed the Arrhenius theory. These findings constructed a prediction model for bifidobacteria survival in powder form.     

The L1/2 regularization method for variable selection in the Cox model

In this paper, we investigate to use the L_1/2 regularization method for variable selection based on the Cox's proportional hazards model. The L_1/2 regularization can be taken as a representative of L_q (0 < q < 1) regularizations and has been demonstrated many attractive properties. To solve the L_1/2 penalized Cox model, we propose a coordinate descent algorithm with a new univariate half thresholding operator which is applicable to high-dimensional biological data. Simulation results based on standard artificial data show that the L_1/2 regularization method can be more accurate for variable selection than Lasso and SCAD methods. The results from real DNA microarray datasets indicate the L_1/2 regularization method performs competitively.     

An intelligent approach to machine component health prognostics by utilizing only truncated histories

Numerous techniques and methods have been proposed to reduce the production downtime, spare-part inventory, maintenance cost, and safety hazards of machineries and equipment. Prognostics are regarded as a significant and promising tool for achieving these benefits for machine maintenance. However, prognostic models, particularly probabilistic-based methods, require a large number of failure instances. In practice, engineering assets are rarely being permitted to run to failure. Many studies have reported valuable models and methods that engage in maximizing both truncated and failure data. However, limited studies have focused on cases where only truncated data are available, which is common in machine condition monitoring. Therefore, this study develops an intelligent machine component prognostics system by utilizing only truncated histories. First, the truncated Minimum Quantization Error (MQE) histories were obtained by Self-organizing Map network after feature extraction. The chaos-based parallel multilayer perceptron network and polynomial fitting for residual errors were adopted to generate the predicted MQEs and failure times following the truncation times. The feed-forward neural network (FFNN) was trained with inputs both from the truncated MQE histories and from the predicted MQEs. The target vectors of survival probabilities were estimated by intelligent product limit estimator using the truncation times and generated failure times. After validation, the FFNN was applied to predict the machine component health of individual units. To validate the proposed method, two cases were considered by using the degradation data generated by bearing testing rig. Results demonstrate that the proposed method is a promising intelligent prognostics approach for machine component health.     

Life history of Trogoderma granarium Everts (Coleoptera: Dermestidae) on peeled barley,peeled oats and triticale

The life history of the khapra beetle, Trogoderma granarium Everts (Coleoptera: Dermestidae), was investigated on cracked peeled barley, cracked peeled oats and cracked triticale. The results suggest that the biological and demographic parameters of T. granarium are affected by the tested commodities. The developmental time of total immature period of T. granarium was significantly lower on barley (35.06 days) than on oats (42.87 days) and triticale (43.58 days). Mean longevities of adult females were 7.00, 7.89 and 8.24 days, and for adult males were 5.79, 6.05 and 5.49 days when T. granarium fed on barley, oats and triticale, respectively. The highest fecundity recorded when T. granarium fed on barley (39.0 eggs/female) or triticale (35.2 eggs/female), but it was significantly lower when this species fed on oats (25.4 eggs/female). While the net reproductive rate did not differ significantly when T. granarium fed on barley (14.6 females/female), oats 10.2 (females/female) and triticale (11.0 females/female), the mean generation time was significantly lower when fed on barley (40.5 days) than on oats (48.7 days) and triticale (46.1 days). The intrinsic rate of increase of T. granarium did not differ significantly when fed on barley and triticale (0.066 and 0.052 females/female/day, respectively), but it was significantly lower when fed on oats (0.047 females/female/day). The biological and demographic parameters of T. granarium may be incorporated in population models that evaluate the population dynamics aiming to the optimization of management strategies against this noxious species.     

Joint latent class model of survival and longitudinal data: An application to CPCRA study

There has been an increasing interest in the joint analysis of repeated measures and time to event data. In many studies, there could also exist heterogeneous subgroups. Thus a new model is proposed for the joint analysis of longitudinal and survival data with underlying subpopulations identified by latent class model. Within each latent class, a joint model of longitudinal and survival data with shared random effects is adopted. The proposed model is applied to Terry Beirn Community Programs for Clinical Research on AIDS study (CPCRA) to characterize the underlying heterogeneity of the cohort and to study the relation between longitudinal CD4 measures and time to death. The proposed model is desirable when the heterogeneity among subjects cannot be ignored and both the longitudinal and survival outcomes are of interest.     

Modeling the inactivation of Neosartorya fischeri ascospores in apple juice by high pressure,power ultrasound and thermal processing

Neosartorya fischeri is a mould that spoils acid foods and can produce mycotoxins. In this work, the efficacy of high pressure processing (HPP, 600 MPa) and power ultrasound (24 kHz, 0.33 W/mL) in combination with 75 °C for the inactivation of four week old N. fischeri ascospores in apple juice was investigated and compared with 75 °C thermal processing alone. The HPP-75 °C process was the most effective technique for inactivating N. fischeri spores, resulting in 3.3 log reductions after 10 min vs. no inactivation for thermosonication (TS) and thermal processing. Unexpectedly, activation shoulders were observed during the TS process. Then, the effect of different temperatures on the ascospore inactivation in apple juice by HPP-thermal, TS and thermal processing was investigated, and the log survivors vs. time were modeled. Faster inactivation was achieved at higher temperatures for all the technologies tested, indicating the significant role of temperature for the spore inactivation, alone or combined with other processes. The Weibull model described the spore inactivation better by 600 MPa HPP-thermal (50, 60, 75 °C) and thermal (85, 90 °C), whereas Lorentzian was more appropriate for the TS treatment (65, 70, 75 °C). In conclusion, HPP is the best food preservation technology due to higher spore inactivation in apple juice at the same temperature.     

Behavioral assessment of recoverable credit of retailer’s customers

The increasing rate of late payments by credit card customers, which are caused by the recent economic downturn, is causing not only reduced profit margins but also significant sales losses for retail companies. Under pressure to increase revenues, credit prediction should be a part of customer delinquency management. In this study, a credit prediction model has been developed to manage delinquents holding retail credit cards. The hybrid model combines a Kohonen network and a Cox’s proportional hazard model. A Kohonen network is used to cluster credit delinquents into homogeneous groups. A Cox’s hazard model is used to analyze repayment patterns of delinquents in each group. The model estimates the expected time of credit recovery from delinquents. This model’s prediction accuracy scored above 93%.