Bankruptcy prediction using fuzzy methods

In this thesis, a classi cation problem in predicting credit worthiness of a customer is tackled. This is done by proposing a reliable classi cation procedure on a given data set. The aim of this thesis is to design a model that gives the best classi cation accuracy to e ectively predict bankruptcy. FRPCA techniques proposed by Yang and Wang have been preferred since they are tolerant to certain type of noise in the data. These include FRPCA1, FRPCA2 and FRPCA3 from which the best method is chosen. Two di erent approaches are used at the classi cation stage: Similarity classi er and FKNN classi er. Algorithms are tested with Australian credit card screening data set. Results obtained indicate a mean classi cation accuracy of 83.22% using FRPCA1 with similarity classi- er. The FKNN approach yields a mean classi cation accuracy of 85.93% when used with FRPCA2, making it a better method for the suitable choices of the number of nearest neighbors and fuzziness parameters. Details on the calibration of the fuzziness parameter and other parameters associated with the similarity classi er are discussed.

Feature selection using Fuzzy Entropy measures with Yu's Similarity measure

In this study, feature selection in classification based problems is highlighted. The role of feature selection methods is to select important features by discarding redundant and irrelevant features in the data set, we investigated this case by using fuzzy entropy measures. We developed fuzzy entropy based feature selection method using Yu's similarity and test this using similarity classifier. As the similarity classifier we used Yu's similarity, we tested our similarity on the real world data set which is dermatological data set. By performing feature selection based on fuzzy entropy measures before classification on our data set the empirical results were very promising, the highest classification accuracy of 98.83% was achieved when testing our similarity measure to the data set. The achieved results were then compared with some other results previously obtained using different similarity classifiers, the obtained results show better accuracy than the one achieved before. The used methods helped to reduce the dimensionality of the used data set, to speed up the computation time of a learning algorithm and therefore have simplified the classification task

Evaluation of follicular lymphoid depletion in the Bursa of Fabricius: an alternative methodology using digital image analysis and artificial neural networks

Fifty Bursa of Fabricius (BF) were examined by conventional optical microscopy and digital images were acquired and processed using Matlab® 6.5 software. The Artificial Neuronal Network (ANN) was generated using Neuroshell® Classifier software and the optical and digital data were compared. The ANN was able to make a comparable classification of digital and optical scores. The use of ANN was able to classify correctly the majority of the follicles, reaching sensibility and specificity of 89% and 96%, respectively. When the follicles were scored and grouped in a binary fashion the sensibility increased to 90% and obtained the maximum value for the specificity of 92%. These results demonstrate that the use of digital image analysis and ANN is a useful tool for the pathological classification of the BF lymphoid depletion. In addition it provides objective results that allow measuring the dimension of the error in the diagnosis and classification therefore making comparison between databases feasible.

Learning Grasp Affordances from Vision

Monimutkaisissa ja muuttuvissa ympäristöissä työskentelevät robotit tarvitsevat kykyä manipuloida ja tarttua esineisiin. Tämä työ tutkii robottitarttumisen ja robottitartuntapis-teiden koneoppimisen aiempaa tutkimusta ja nykytilaa. Nykyaikaiset menetelmät käydään läpi, ja Le:n koneoppimiseen pohjautuva luokitin toteutetaan, koska se tarjoaa parhaan onnistumisprosentin tutkituista menetelmistä ja on muokattavissa sopivaksi käytettävissä olevalle robotille. Toteutettu menetelmä käyttää intensititeettikuvaan ja syvyyskuvaan po-hjautuvia ominaisuuksi luokitellakseen potentiaaliset tartuntapisteet. Tämän toteutuksen tulokset esitellään.

Variability in the fecal egg count and the parasitic burden of hair sheep after grazing in nematode infected paddocks

This study aimed to evaluate the variability in the fecal egg count (FEC) and the parasitic burden of naive hair sheep after grazing in nematode infected paddocks. The research was carried out in Tabasco, Mexico, during two periods (August and December). In each period 32 lambs were grazed for one month on African star grass (Cynodon plectostachyus) contaminated with gastrointestinal parasitic nematodes. FEC, packed cell volume (PCV) and body weight (BW) were recorded. Gastrointestinal worms were recovered at necropsy. Data were analyzed with the MIXED procedure of SAS using a model of repeated measurements over time. A higher number of Haemonchus contortus worms was found in December (2814±838) than in August (1166±305). The opposite occurred with Cooperia curticei (2167±393 and 3638±441, respectively). The FEC and correlation coefficient in respect to the worm burden were higher in December (6516 ± 1599, r=0.83, respectively) than in August (4364±771, r=0.44, respectively). A high variability in resistance-susceptibility to gastrointestinal nematodes (GIN) occurred in Katahdin × Pelibuey lambs after grazing.

Additional information about a mange outbreak by Allopsoroptoides galli (Acari: Psoroptoididae) in commercial laying hens in the state of São Paulo, Brazil

This paper reports additional information about a mange outbreak by the mite Allopsoroptoides galli in a commercial egg-laying hen facility in the state of São Paulo, Brazil. About half of the 76,000 multi-age birds of the flock were affected. Experimental infestations carried out on naive hens resulted in clinical signs similar to those diagnosed in naturally infested hens, such as generalized scaly dermatitis, presence of mucus-like material and yellowish crusts on the skin and around the calami, feather loss and strong unpleasant odor. About 30% drop of egg production was estimated. The possible source of infestation were wild birds identified on the ground and roofs of the sheds.

Steganografia ja kuvien digitaalinen vesileimaus

Steganografian tarkoituksena on salaisen viestin piilottaminen muun informaation sekaan. Tutkielmassa perehdytään kirjallisuuden pohjalta steganografiaan ja kuvien digitaaliseen vesileimaamiseen. Tutkielmaan kuuluu myös kokeellinen osuus. Siinä esitellään vesileimattujen kuvien tunnistamiseen kehitetty testausjärjestelmä ja testiajojen tulokset. Testiajoissa kuvasarjoja on vesileimattu valituilla vesileimausmenetelmillä parametreja vaihdellen. Tunnistettaville kuville tehdään piirreirrotus. Erotellut piirteet annetaan parametreina luokittimelle, joka tekee lopullisen tunnistamispäätöksen. Tutkimuksessa saatiin toteutettua toimiva ohjelmisto vesileiman lisäämiseen ja vesileimattujen kuvien tunnistamiseen kuvajoukosta. Tulosten perusteella, sopivalla piirreirrottimella ja tukivektorikoneluokittimella päästään yli 95 prosentin tunnistamistarkkuuteen.

Oral tolerance induction with altered forms of ovalbumin

As a T cell-dependent phenomenon, oral tolerance is not expected to depend necessarily on native configuration of antigens. We investigated the induction of oral tolerance with modified ovalbumin (Ova). Oral administration of heat-denatured (HD-Ova) and cyanogen bromide-degraded ovalbumin was less effective than native Ova in inducing oral tolerance in B6D2F1 mice. HD-Ova was effective in suppressing delayed-type hypersensitivity (DTH) reactions but did not suppress specific antibody formation. Injection of Ova directly into the stomach, but not into the ileum or cecum, suppressed subsequent immunization to DTH reactions. Gavage with protease inhibitors (aprotinin or ovomucoid) before gavage with Ova was ineffective in blocking tolerance induction. Treatment with hydroxyurea to destroy cycling cells 24 h before gavage with Ova blocked oral tolerance induction and also the possibility to passively transfer tolerance to naive recipients with the serum of mice gavaged with Ova 1 h before. The implications of these findings about oral tolerance induction are discussed

Neuroethologic differences in sleep deprivation induced by the single- and multiple-platform methods

It has been proposed that the multiple-platform method (MP) for desynchronized sleep (DS) deprivation eliminates the stress induced by social isolation and by the restriction of locomotion in the single-platform (SP) method. MP, however, induces a higher increase in plasma corticosterone and ACTH levels than SP. Since deprivation is of heuristic value to identify the functional role of this state of sleep, the objective of the present study was to determine the behavioral differences exhibited by rats during sleep deprivation induced by these two methods. All behavioral patterns exhibited by a group of 7 albino male Wistar rats submitted to 4 days of sleep deprivation by the MP method (15 platforms, spaced 150 mm apart) and by 7 other rats submitted to sleep deprivation by the SP method were recorded in order to elaborate an ethogram. The behavioral patterns were quantitated in 10 replications by naive observers using other groups of 7 rats each submitted to the same deprivation schedule. Each quantification session lasted 35 min and the behavioral patterns presented by each rat over a period of 5 min were counted. The results obtained were: a) rats submitted to the MP method changed platforms at a mean rate of 2.62 ± 1.17 platforms h-1 animal-1; b) the number of episodes of noninteractive waking patterns for the MP animals was significantly higher than that for SP animals (1077 vs 768); c) additional episodes of waking patterns (26.9 ± 18.9 episodes/session) were promoted by social interaction in MP animals; d) the cumulative number of sleep episodes observed in the MP test (311) was significantly lower (chi-square test, 1 d.f., P<0.05) than that observed in the SP test (534); e) rats submitted to the MP test did not show the well-known increase in ambulatory activity observed after the end of the SP test; f) comparison of 6 MP and 6 SP rats showed a significantly shorter latency to the onset of DS in MP rats (7.8 ± 4.3 and 29.0 ± 25.0 min, respectively; Student t-test, P<0.05). We conclude that the social interaction occurring in the MP test generates additional stress since it increases the time of forced wakefulness and reduces the time of rest promoted by synchronized sleep.

Experimental anxiety and the reinforcing effects of ethanol in rats

In order to examine the relationship between anxiety and reinforcing effects of alcohol, drug-naive male Wistar rats weighing 250-300 g were classified as "anxious" and "non-anxious" in the elevated plus-maze test. A conditioned place preference test was then used to investigate the reinforcing effects of ethanol (EtOH) on these animals. On 2 alternate days, groups of "anxious", "non-anxious" and "normal" rats received intraperitoneal (ip) injections of EtOH (0.5, 1.0 or 1.5 g/kg) immediately before a 15-min confinement to the white compartment. On the 2 intervening days the same rats received ip injections of saline before confinement to the opposite compartment. On day 5, a 15-min free-choice test was carried out with no injections. Rats classified as "anxious" showed a significant, though not dose-dependent preference for all doses of ethanol compared to saline-treated animals. These data demonstrate that rats regarded as "anxious" are more sensitive to the reinforcing effects of EtOH than "non-anxious" and "normal" Wistar rats and emphasize the relevance of the basal levels of anxiety of rats when trying to detect the reinforcing effects of EtOH.

Ischemic preconditioning reduces peripheral oxidative damage associated with brain ischemia in rats

Brain ischemia followed by reperfusion causes neuronal death related to oxidative damage. Furthermore, it has been reported that subjects suffering from ischemic cerebrovascular disorders exhibit changes in circulating platelet aggregation, a characteristic that might be important for their clinical outcome. In the present investigation we studied tert-butyl hydroperoxide-initiated plasma chemiluminescence and thiol content as measures of peripheral oxidative damage in naive and preconditioned rats submitted to forebrain ischemia produced by the 4-vessel occlusion method. Rats were submitted to 2 or 10 min of global transient forebrain ischemia followed by 60 min or 1, 2, 5, 10 or 30 days of reperfusion. Preconditioned rats were submitted to a 10-min ischemic episode 1 day after a 2-min ischemic event (2 + 10 min), followed by 60 min or 1 or 2 days of reperfusion. It has been demonstrated that such preconditioning protects against neuronal death in rats and gerbils submitted to a lethal (10 min) ischemic episode. The results show that both 2 and 10 min of ischemia cause an increase of plasma chemiluminescence when compared to control and sham rats. In the 2-min ischemic group, the effect was not present after reperfusion. In the 10-min ischemic group, the increase was present up to 1 day after recirculation and values returned to control levels after 2 days. However, rats preconditioned to ischemia (2 + 10 min) and reperfusion showed no differences in plasma chemiluminescence when compared to controls. We also analyzed plasma thiol content since it has been described that sulfhydryl (SH) groups significantly contribute to the antioxidant capacity of plasma. There was a significant decrease of plasma thiol content after 2, 10 and 2 + 10 min of ischemia followed by reperfusion when compared to controls. We conclude that ischemia may cause, along with brain oxidative damage and cell death, a peripheral oxidative damage that is reduced by the preconditioning phenomenon.

Chronic intrathecal cannulation enhances nociceptive responses in rats

The influence of a chronically implanted spinal cannula on the nociceptive response induced by mechanical, chemical or thermal stimuli was evaluated. The hyperalgesia in response to mechanical stimulation induced by carrageenin or prostaglandin E2 (PGE2) was significantly increased in cannulated (Cn) rats, compared with naive (Nv) or sham-operated (Sh) rats. Only Cn animals presented an enhanced nociceptive response in the first phase of the formalin test when low doses were used (0.3 and 1%). The withdrawal latency to thermal stimulation of a paw inflamed by carrageenin was significantly reduced in Cn rats but not in Nv or Sh rats. In contrast to Nv and Sh rats, injection in Cn animals of a standard non-steroid anti-inflammatory drug, indomethacin, either intraperitoneally or into the spinal cord via an implanted cannula or by direct puncture of the intrathecal space significantly blocked the intensity of the hyperalgesia induced by PGE2. Cannulated animals treated with indomethacin also showed a significant inhibition of second phase formalin-induced paw flinches. Histopathological analysis of the spinal cord showed an increased frequency of mononuclear inflammatory cells in the Cn groups. Thus, the presence of a chronically implanted cannula seems to cause nociceptive spinal sensitization to mechanical, chemical and thermal stimulation, which can be blocked by indomethacin, thus suggesting that it may result from the spinal release of prostaglandins due to an ongoing mild inflammation.

Linguistic models for decision support

Linguistic modelling is a rather new branch of mathematics that is still undergoing rapid development. It is closely related to fuzzy set theory and fuzzy logic, but knowledge and experience from other fields of mathematics, as well as other fields of science including linguistics and behavioral sciences, is also necessary to build appropriate mathematical models. This topic has received considerable attention as it provides tools for mathematical representation of the most common means of human communication - natural language. Adding a natural language level to mathematical models can provide an interface between the mathematical representation of the modelled system and the user of the model - one that is sufficiently easy to use and understand, but yet conveys all the information necessary to avoid misinterpretations. It is, however, not a trivial task and the link between the linguistic and computational level of such models has to be established and maintained properly during the whole modelling process. In this thesis, we focus on the relationship between the linguistic and the mathematical level of decision support models. We discuss several important issues concerning the mathematical representation of meaning of linguistic expressions, their transformation into the language of mathematics and the retranslation of mathematical outputs back into natural language. In the first part of the thesis, our view of the linguistic modelling for decision support is presented and the main guidelines for building linguistic models for real-life decision support that are the basis of our modeling methodology are outlined. From the theoretical point of view, the issues of representation of meaning of linguistic terms, computations with these representations and the retranslation process back into the linguistic level (linguistic approximation) are studied in this part of the thesis. We focus on the reasonability of operations with the meanings of linguistic terms, the correspondence of the linguistic and mathematical level of the models and on proper presentation of appropriate outputs. We also discuss several issues concerning the ethical aspects of decision support - particularly the loss of meaning due to the transformation of mathematical outputs into natural language and the issue or responsibility for the final decisions. In the second part several case studies of real-life problems are presented. These provide background and necessary context and motivation for the mathematical results and models presented in this part. A linguistic decision support model for disaster management is presented here – formulated as a fuzzy linear programming problem and a heuristic solution to it is proposed. Uncertainty of outputs, expert knowledge concerning disaster response practice and the necessity of obtaining outputs that are easy to interpret (and available in very short time) are reflected in the design of the model. Saaty’s analytic hierarchy process (AHP) is considered in two case studies - first in the context of the evaluation of works of art, where a weak consistency condition is introduced and an adaptation of AHP for large matrices of preference intensities is presented. The second AHP case-study deals with the fuzzified version of AHP and its use for evaluation purposes – particularly the integration of peer-review into the evaluation of R&D outputs is considered. In the context of HR management, we present a fuzzy rule based evaluation model (academic faculty evaluation is considered) constructed to provide outputs that do not require linguistic approximation and are easily transformed into graphical information. This is achieved by designing a specific form of fuzzy inference. Finally the last case study is from the area of humanities - psychological diagnostics is considered and a linguistic fuzzy model for the interpretation of outputs of multidimensional questionnaires is suggested. The issue of the quality of data in mathematical classification models is also studied here. A modification of the receiver operating characteristics (ROC) method is presented to reflect variable quality of data instances in the validation set during classifier performance assessment. Twelve publications on which the author participated are appended as a third part of this thesis. These summarize the mathematical results and provide a closer insight into the issues of the practicalapplications that are considered in the second part of the thesis.

Rapid eye movement sleep deprivation induces an increase in acetylcholinesterase activity in discrete rat brain regions

Some upper brainstem cholinergic neurons (pedunculopontine and laterodorsal tegmental nuclei) are involved in the generation of rapid eye movement (REM) sleep and project rostrally to the thalamus and caudally to the medulla oblongata. A previous report showed that 96 h of REM sleep deprivation in rats induced an increase in the activity of brainstem acetylcholinesterase (Achase), the enzyme which inactivates acetylcholine (Ach) in the synaptic cleft. There was no change in the enzyme's activity in the whole brain and cerebrum. The components of the cholinergic synaptic endings (for example, Achase) are not uniformly distributed throughout the discrete regions of the brain. In order to detect possible regional changes we measured Achase activity in several discrete rat brain regions (medulla oblongata, pons, thalamus, striatum, hippocampus and cerebral cortex) after 96 h of REM sleep deprivation. Naive adult male Wistar rats were deprived of REM sleep using the flower-pot technique, while control rats were left in their home cages. Total, membrane-bound and soluble Achase activities (nmol of thiocholine formed min-1 mg protein-1) were assayed photometrically. The results (mean ± SD) obtained showed a statistically significant (Student t-test) increase in total Achase activity in the pons (control: 147.8 ± 12.8, REM sleep-deprived: 169.3 ± 17.4, N = 6 for both groups, P<0.025) and thalamus (control: 167.4 ± 29.0, REM sleep-deprived: 191.9 ± 15.4, N = 6 for both groups, P<0.05). Increases in membrane-bound Achase activity in the pons (control: 171.0 ± 14.7, REM sleep-deprived: 189.5 ± 19.5, N = 6 for both groups, P<0.05) and soluble enzyme activity in the medulla oblongata (control: 147.6 ± 16.3, REM sleep-deprived: 163.8 ± 8.3, N = 6 for both groups, P<0.05) were also observed. There were no statistically significant differences in the enzyme's activity in the other brain regions assayed. The present findings show that the increase in Achase activity induced by REM sleep deprivation was specific to the pons, a brain region where cholinergic neurons involved in REM generation are located, and also to brain regions which receive cholinergic input from the pons (the thalamus and medulla oblongata). During REM sleep extracellular levels of Ach are higher in the pons, medulla oblongata and thalamus. The increase in Achase activity in these brain areas after REM sleep deprivation suggests a higher rate of Ach turnover.