Structured output SVM for remote sensing image classification

In the recent years, kernel methods have revealed very powerful tools in many application domains in general and in remote sensing image classification in particular. The special characteristics of remote sensing images (high dimension, few labeled samples and different noise sources) are efficiently dealt with kernel machines. In this paper, we propose the use of structured output learning to improve remote sensing image classification based on kernels. Structured output learning is concerned with the design of machine learning algorithms that not only implement input-output mapping, but also take into account the relations between output labels, thus generalizing unstructured kernel methods. We analyze the framework and introduce it to the remote sensing community. Output similarity is here encoded into SVM classifiers by modifying the model loss function and the kernel function either independently or jointly. Experiments on a very high resolution (VHR) image classification problem shows promising results and opens a wide field of research with structured output kernel methods.

Generalized descriptive set theory and classification theory

Descriptive set theory is mainly concerned with studying subsets of the space of all countable binary sequences. In this paper we study the generalization where countable is replaced by uncountable. We explore properties of generalized Baire and Cantor spaces, equivalence relations and their Borel reducibility. The study shows that the descriptive set theory looks very different in this generalized setting compared to the classical, countable case. We also draw the connection between the stability theoretic complexity of first-order theories and the descriptive set theoretic complexity of their isomorphism relations. Our results suggest that Borel reducibility on uncountable structures is a model theoretically natural way to compare the complexity of isomorphism relations.

Age effects, unobserved characteristics and hedonic price indexes: The Spanish car market in the 1990's

This paper computes and compares alternative quality-adjusted price indexes for new cars in Spain in the period 1990-2000. The proposed hedonic approach simultaneously controls for time-invariant unobserved product effects and time-variant unobserved quality changes, that are assumed to be captured by model age effects. The results show that the non-adjusted price index largely overstates the increase in the cost of living induced by changes in car prices and that previous evidence for this market have not measured the real extent of that bias, probably due to the omission of controls for unobservables. It is also shown that omitting age effects can also lead to misleading conclusions. The estimated price indexes give also some insights on what could have been the determinants of price evolution in the Spanish car market. JEL classi…fication numbers: C43, E31, L11, L13, Keywords: Hedonic price indexes, Spanish car market, car prices, CPI, Cost of living

Numerical Taxonomy of Old World Phlebotominae (Diptera: Psychodidae): 2. Restatement of Classification upon Subgeneric Morphological Characters

Numerical analyses (correspondence analysis, ascending hierarchical classification, and cladistics) were done with morphological characters of adult phlebotomine sand flies. The resulting classification largely confirms that of classical taxonomy for supra-specific groups from the Old World, though the positions of some groups are adjusted. The taxa Spelaeophlebotomus Theodor 1948, Idiophlebotomus Quate & Fairchild 1961, Australophlebotomus Theodor 1948 and Chinius Leng 1987 are notably distinct from other Old World groups, particularly from the genus Phlebotomus Rondani & Berté 1840. Spelaeomyia Theodor 1948 and, in particular, Parvidens Theodor & Mesghali 1964 are clearly separate from Sergentomyia França & Parrot 1920.

Rapidly fatal poisoning with an insecticide containing rotenone.

Introduction: Rotenone is a botanical pesticide derived from extracts of Derris roots, which is traditionally used as piscicide, but also as an industrial insecticide for home gardens. Its mechanism of action is potent inhibition of mitochondrial respiratory chain by uncoupling oxidative phosphorylation by blocking electron transport at complex-I. Despite its classification as mild to moderately toxic to humans (estimated LD50, 300-500 mg/kg), there is a striking variety of acute toxicity of rotenone depending on the formulation (solvents). Human fatalities with rotenone-containing insecticides have been rarely reported, and a rapid deterioration within a few hours of the ingestion has been described previously in one case. Case report: A 49-year-old Tamil man with a history of asthma, ingested 250 mL of an insecticide containing 1.24% of rotenone (3.125 g, 52.1-62.5 mg/kg) in a suicide attempt at home. The product was not labeled as toxic. One hour later, he vomited repeatedly and emergency services were alerted. He was found unconscious with irregular respiration and was intubated. On arrival at the emergency department, he was comatose (GCS 3) with fixed and dilated pupils, and absent corneal reflexes. Physical examination revealed hemodynamic instability with hypotension (55/30 mmHg) and bradycardia (52 bpm). Significant laboratory findings were lactic acidosis (pH 6.97, lactate 17 mmol/L) and hypokalemia (2 mmol/L). Cranial computed tomography (CT) showed early cerebral edema. A single dose of activated charcoal was given. Intravenous hydration, ephedrine, repeated boli of dobutamine, and a perfusor with 90 micrograms/h norepinephine stabilized blood pressure temporarily. Atropine had a minimal effect on heart rate (58 bpm). Intravenous lipid emulsion was considered (log Pow 4.1), but there was a rapid deterioration with refractory hypotension and acute circulatory failure. The patient died 5h after ingestion of the insecticide. No autopsy was performed. Quantitative analysis of serum performed by high-resolution/accurate mass-mass spectrometry and liquid chromatography (LC-HR/AM-MS): 560 ng/mL rotenone. Other substances were excluded by gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS/MS). Conclusion: The clinical course was characterized by early severe symptoms and a rapidly fatal evolution, compatible with inhibition of mitochondrial energy supply. Although rotenone is classified as mild to moderately toxic, physicians must be aware that suicidal ingestion of emulsified concentrates may be rapidly fatal. (n=3): stridor, cyanosis, cough (one each). Local swelling after chewing or swallowing soap developed at the earliest after 20 minutes and persisted beyond 24 hours in some cases. Treatment with antihistamines and/or steroids relieved the symptoms in 9 cases. Conclusion: Bar soap ingestion by seniors carries a risk of severe local reactions. Half the patients developed symptoms, predominantly swellings of tongue and/or lips (38%). Cognitive impairment, particularly in the cases of dementia (37%), may increase the risk of unintentional ingestion. Chewing and intraoral retention of soap leads to prolonged contact with the mucosal membranes. Age-associated physiological changes of oral mucosa probably promote the irritant effects of the surfactants. Medical treatment with antihistamines and corticosteroids usually leads to rapid decline of symptoms. Without treatment, there may be a risk of airway obstruction.

Textural classification of land cover using support vector machines: an empirical comparison with parametric, non parametric and hybrid classifiers in the Bolivian Amazon

Land cover classification is a key research field in remote sensing and land change science as thematic maps derived from remotely sensed data have become the basis for analyzing many socio-ecological issues. However, land cover classification remains a difficult task and it is especially challenging in heterogeneous tropical landscapes where nonetheless such maps are of great importance. The present study aims to establish an efficient classification approach to accurately map all broad land cover classes in a large, heterogeneous tropical area of Bolivia, as a basis for further studies (e.g., land cover-land use change). Specifically, we compare the performance of parametric (maximum likelihood), non-parametric (k-nearest neighbour and four different support vector machines - SVM), and hybrid classifiers, using both hard and soft (fuzzy) accuracy assessments. In addition, we test whether the inclusion of a textural index (homogeneity) in the classifications improves their performance. We classified Landsat imagery for two dates corresponding to dry and wet seasons and found that non-parametric, and particularly SVM classifiers, outperformed both parametric and hybrid classifiers. We also found that the use of the homogeneity index along with reflectance bands significantly increased the overall accuracy of all the classifications, but particularly of SVM algorithms. We observed that improvements in producer’s and user’s accuracies through the inclusion of the homogeneity index were different depending on land cover classes. Earlygrowth/degraded forests, pastures, grasslands and savanna were the classes most improved, especially with the SVM radial basis function and SVM sigmoid classifiers, though with both classifiers all land cover classes were mapped with producer’s and user’s accuracies of around 90%. Our approach seems very well suited to accurately map land cover in tropical regions, thus having the potential to contribute to conservation initiatives, climate change mitigation schemes such as REDD+, and rural development policies.

Innovation or Imitation? The effect of spillovers and competitive pressure on firms' R&D strategy choice

This article provides a theoretical and empirical analysis of a firm's optimal R&D strategy choice. In this paper a firm's R&D strategy is assumed to be endogenous and allowed to depend on both internal firms. characteristics and external factors. Firms choose between two strategies, either they engage in R&D or abstain from own R&D and imitate the outcomes of innovators. In the theoretical model this yields three types of equilibria in which either all firms innovate, some firms innovate and others imitate, or no firm innovates. Firms'equilibrium strategies crucially depend on external factors. We find that the efficiency of intellectual property rights protection positively affects firms'incentives to engage in R&D, while competitive pressure has a negative effect. In addition, smaller firms are found to be more likely to become imitators when the product is homogeneous and the level of spillovers is high. These results are supported by empirical evidence for German .rms from manufacturing and services sectors. Regarding social welfare our results indicate that strengthening intellectual property protection can have an ambiguous effect. In markets characterized by a high rate of innovation a reduction of intellectual property rights protection can discourage innovative performance substantially. However, a reduction of patent protection can also increase social welfare because it may induce imitation. This indicates that policy issues such as the optimal length and breadth of patent protection cannot be resolved without taking into account specific market and firm characteristics. Journal of Economic Literature Classification Numbers: C35, D43, L13, L22, O31. Keywords: Innovation; imitation; spillovers; product differentiation; market competition; intellectual property rights protection.

Profitability, uncertainty and multi-product firm product proliferation: The Spanish car industry

This article studies how product introduction decisions relate to profitability and uncertainty in the context of multi-product firms and product differentiation. These two features, common to many modern industries, have not received much attention in the literature as compared to the classical problem of firm entry, even if the determinants of firm and product entry are quite different. The theoretical predictions about the sign of the impact of uncertainty on product entry are not conclusive. Therefore, an econometric model relating firms’ product introduction decisions with profitability and profit uncertainty is proposed. Firm’s estimated profits are obtained from a structural model of product demand and supply, and uncertainty is proxied by profits’ variance. The empirical analysis is carried out using data on the Spanish car industry for the period 1990-2000. The results show a positive relationship between product introduction and profitability, and a negative one with respect to profit variability. Interestingly, the degree of uncertainty appears to be a driving force of entry stronger than profitability, suggesting that the product proliferation process in the Spanish car market may have been mainly a consequence of lower uncertainty rather than the result of having a more profitable market. Keywords: Product introduction, entry, uncertainty, multiproduct firms, automobile JEL codes: L11, L13

International consensus conference on PFAPA syndrome: evaluation of a new set of classification criteria

INTRODUCTION: PFAPA syndrome is characterized by periodic fever, associated with pharyngitis, cervical adenitis and/or aphthous stomatitis and belongs to the auto-inflammatory diseases. Diagnostic criteria are based on clinical features and the exclusion of other periodic fever syndromes. An analysis of a large cohort of patients has shown weaknesses for these criteria and there is a lack of international consensus. An International Conference was held in Morges in November 2008 to propose a new set of classification criteria based on a consensus among experts in the field.OBJECTIVE: We aimed to verify the applicability of the new set of classification criteria.PATIENTS & METHODS: 80 patients diagnosed with PFAPA syndrome from 3 centers (Genoa, Lausanne and Geneva) for pediatric rheumatology were included in the study. A detailed description of the clinical and laboratory features was obtained. The new classification criteria and the actual diagnostic criteria were applied to the patients.RESULTS: Only 40/80 patients (50%) fulfilled all criteria of the new classification. 31 patients were excluded because they didn't meet one of the 7 diagnostic criteria, 7 because of 2 criteria, and one because of 3 criteria. When we applied the current criteria to the same patients, 11/80 patients (13.7%) needed to be excluded. 8/80 patients (10%) were excluded from both sets. Exclusion was related only to some of the criteria. Number of patients for each not fulfilled criterion (new set of criteria/actual criteria): age (1/6), symptoms between episodes (2/2), delayed growth (4/1), main symptoms (21/0), periodicity, length of fever, interval between episodes, and length of disease (20/0). The application of some of the new criteria was not easy, as they were both very restrictive and needed precise information from the patients.CONCLUSION: Our work has shown that the new set of classification criteria can be applied to patients suspected for PFAPA syndrome, but it seems to be more restrictive than the actual diagnostic criteria. A further work of validation needs to be done in order to determine if this new set of classification criteria allow a good discrimination between PFAPA patients and other causes of recurrent fever syndromes.

Dynamic product diversity

The goal of this paper is to study the frequency of new product introductions in monopoly markets where demand is subject to transitory saturation. We focus on those types of goods for which consumers purchase at most one unit of each variety, but repeat purchases in the same product category. The model considers infinitely-lived, forward-looking consumers and firms. We show that the share of potential surplus that a monopolist is able to appropriate increases with the frequency of introduction of new products and the intensity of transitory saturation. If the latter is sufficiently strong then the rate of introduction of new products is higher than socially desirable (excessive dynamic product diversity.)