Reservoir Operation for Hydropower Optimization: A Chance-Constrained Approach

This paper presents a chance-constrained linear programming formulation for reservoir operation of a multipurpose reservoir. The release policy is defined by a chance constraint that the probability of irrigation release in any period equalling or exceeding the irrigation demand is at least equal to a specified value P (called reliability level). The model determines the maximum annual hydropower produced while meeting the irrigation demand at a specified reliability level. The model considers variation in reservoir water level elevation and also the operating range within which the turbine operates. A linear approximation for nonlinear power production function is assumed and the solution obtained within a specified tolerance limit. The inflow into the reservoir is considered random. The chance constraint is converted into its deterministic equivalent using a linear decision rule and inflow probability distribution. The model application is demonstrated through a case study.

Interval Data Classification under Partial Information: A Chance-Constraint Approach

This paper presents a Chance-constraint Programming approach for constructing maximum-margin classifiers which are robust to interval-valued uncertainty in training examples. The methodology ensures that uncertain examples are classified correctly with high probability by employing chance-constraints. The main contribution of the paper is to pose the resultant optimization problem as a Second Order Cone Program by using large deviation inequalities, due to Bernstein. Apart from support and mean of the uncertain examples these Bernstein based relaxations make no further assumptions on the underlying uncertainty. Classifiers built using the proposed approach are less conservative, yield higher margins and hence are expected to generalize better than existing methods. Experimental results on synthetic and real-world datasets show that the proposed classifiers are better equipped to handle interval-valued uncertainty than state-of-the-art.

Chance constrained uncertain classification via robust optimization

This paper studies the problem of constructing robust classifiers when the training is plagued with uncertainty. The problem is posed as a Chance-Constrained Program (CCP) which ensures that the uncertain data points are classified correctly with high probability. Unfortunately such a CCP turns out to be intractable. The key novelty is in employing Bernstein bounding schemes to relax the CCP as a convex second order cone program whose solution is guaranteed to satisfy the probabilistic constraint. Prior to this work, only the Chebyshev based relaxations were exploited in learning algorithms. Bernstein bounds employ richer partial information and hence can be far less conservative than Chebyshev bounds. Due to this efficient modeling of uncertainty, the resulting classifiers achieve higher classification margins and hence better generalization. Methodologies for classifying uncertain test data points and error measures for evaluating classifiers robust to uncertain data are discussed. Experimental results on synthetic and real-world datasets show that the proposed classifiers are better equipped to handle data uncertainty and outperform state-of-the-art in many cases.

Immunization of male rabbits with sheep luteal receptor to LH results in production of antibodies exhibiting hormone-agonistic and -antagonistic activities

Antibodies to LH/chorionic gonadotrophin receptor (LH/CG-R; molecular weight 67 000), isolated in a homogenous state (established by SDS-PAGE and ligand blotting) from sheep luteal membrane using human CG (hCG)-Sepharose affinity chromatography, were raised in three adult male rabbits (R-I, R-II and R-III). Each of the rabbits received 20-30 mu g oi the purified receptor in Freund's complete adjuvant at a time. Primary immunization was followed by booster injection at intervals. Production of receptor antibodies was monitored by (1) determining the dilution of the serum (IgG fraction) that could specifically bind 50% of I-125-LH/CG-R added and (2) analysing sera for any chance in testosterone levels. Following primary immunization and the first booster, all three rabbits exhibited a 2.5- to 6.0-fold increase in serum testosterone over basal levels and this effect was spread over a period of time (similar to 40 days) coinciding with the rise and fall of receptor antibodies. The maximal antibody titre (ED(50)) produced at this time ranged from 1:350 to 1:100 to below detectable limits for R-I, R-II and R-III respectively. Subsequent immunizations followed by the second booster resulted in a substantial increase in antibody titre (ED(50) of 1:5000) in R-I, but this was not accompanied by any change in serum testosterone over preimmune levels, suggesting that with the progress of immunization the character of the antibody produced had also changed. Two pools of antisera from R-I collected 10 days following the booster (at day 70 (bleed I) and day 290 (bleed II)) were used in further experiments. IgG isolated from bleed I but not from bleed II antiserum showed a dose-dependent stimulation of testosterone production by mouse Leydig cells in vitro, thus confirming the in vivo hormone-mimicking activity antibodies generated during the early immunization phase. The IgG fractions from both bleeds were, however, capable of inhibiting (1) I-125-hCG binding to crude sheep luteal membrane (EC(50) of 1:70 and 1:350 for bleed I and II antisera respectively) and (2) ovine LH-stimulated testosterone production by mouse Leydig cells in vitro, indicating the presence oi antagonistic antibodies irrespective of the period of time during which the rabbits were immunized. The: fact that bleed I-stimulated testosterone production could be inhibited in a dose-dependent manner by the addition of IgG from bleed II to the mouse Leydig cell in vitro assay system showed that the agonistic activity is intrinsic to the bleed I antibody. The receptor antibody (bleed II) was also capable of blocking LH action in vivo, as rabbits passively (for 24 h with LH/CG-R antiserum) as well as actively (for 130 days) immunized against LH/CG-R failed to respond to a bolus injection of LH (50 mu g). At no time, however, was the serum testosterone reduced below the basal level. This study clearly shows that, unlike with LH antibody, attempts to achieve an LH deficiency effect in vivo by resorting to immunization with hole LH receptor is difficult, as receptor antibodies exhibit both hormone-mimicking (agonistic) as well as hormone-blocking (antagonistic) activities.

PocketMatch: A new algorithm to compare binding sites in protein structures

Recognizing similarities and deriving relationships among protein molecules is a fundamental requirement in present-day biology. Similarities can be present at various levels which can be detected through comparison of protein sequences or their structural folds. In some cases similarities obscure at these levels could be present merely in the substructures at their binding sites. Inferring functional similarities between protein molecules by comparing their binding sites is still largely exploratory and not as yet a routine protocol. One of the main reasons for this is the limitation in the choice of appropriate analytical tools that can compare binding sites with high sensitivity. To benefit from the enormous amount of structural data that is being rapidly accumulated, it is essential to have high throughput tools that enable large scale binding site comparison. Results: Here we present a new algorithm PocketMatch for comparison of binding sites in a frame invariant manner. Each binding site is represented by 90 lists of sorted distances capturing shape and chemical nature of the site. The sorted arrays are then aligned using an incremental alignment method and scored to obtain PMScores for pairs of sites. A comprehensive sensitivity analysis and an extensive validation of the algorithm have been carried out. A comparison with other site matching algorithms is also presented. Perturbation studies where the geometry of a given site was retained but the residue types were changed randomly, indicated that chance similarities were virtually non-existent. Our analysis also demonstrates that shape information alone is insufficient to discriminate between diverse binding sites, unless combined with chemical nature of amino acids. Conclusion: A new algorithm has been developed to compare binding sites in accurate, efficient and high-throughput manner. Though the representation used is conceptually simplistic, we demonstrate that along with the new alignment strategy used, it is sufficient to enable binding comparison with high sensitivity. Novel methodology has also been presented for validating the algorithm for accuracy and sensitivity with respect to geometry and chemical nature of the site. The method is also fast and takes about 1/250(th) second for one comparison on a single processor. A parallel version on BlueGene has also been implemented.

Software simulation of the EEG

The literature contains many examples of digital procedures for the analytical treatment of electroencephalograms, but there is as yet no standard by which those techniques may be judged or compared. This paper proposes one method of generating an EEG, based on a computer program for Zetterberg's simulation. It is assumed that the statistical properties of an EEG may be represented by stationary processes having rational transfer functions and achieved by a system of software fillers and random number generators.The model represents neither the neurological mechanism response for generating the EEG, nor any particular type of EEG record; transient phenomena such as spikes, sharp waves and alpha bursts also are excluded. The basis of the program is a valid ‘partial’ statistical description of the EEG; that description is then used to produce a digital representation of a signal which if plotted sequentially, might or might not by chance resemble an EEG, that is unimportant. What is important is that the statistical properties of the series remain those of a real EEG; it is in this sense that the output is a simulation of the EEG. There is considerable flexibility in the form of the output, i.e. its alpha, beta and delta content, which may be selected by the user, the same selected parameters always producing the same statistical output. The filtered outputs from the random number sequences may be scaled to provide realistic power distributions in the accepted EEG frequency bands and then summed to create a digital output signal, the ‘stationary EEG’. It is suggested that the simulator might act as a test input to digital analytical techniques for the EEG, a simulator which would enable at least a substantial part of those techniques to be compared and assessed in an objective manner. The equations necessary to implement the model are given. The program has been run on a DEC1090 computer but is suitable for any microcomputer having more than 32 kBytes of memory; the execution time required to generate a 25 s simulated EEG is in the region of 15 s.

Improved switching frequency variation control of hysteresis controlled voltage source inverter-fed IM drives using current error space vector

A constant switching frequency current error space vector-based hysteresis controller for two-level voltage source inverter-fed induction motor (IM) drives is proposed in this study. The proposed controller is capable of driving the IM in the entire speed range extending to the six-step mode. The proposed controller uses the parabolic boundary, reported earlier, for vector selection in a sector, but uses simple, fast and self-adaptive sector identification logic for sector change detection in the entire modulation range. This new scheme detects the sector change using the change in direction of current error along the axes jA, jB and jC. Most of the previous schemes use an outer boundary for sector change detection. So the current error goes outside the boundary six times during sector change, in one cycle,, introducing additional fifth and seventh harmonic components in phase current. This may cause sixth harmonic torque pulsations in the motor and spread in the harmonic spectrum of phase voltage. The proposed new scheme detects the sector change fast and accurately eliminating the chance of introducing additional fifth and seventh harmonic components in phase current and provides harmonic spectrum of phase voltage, which exactly matches with that of constant switching frequency voltage-controlled space vector pulse width modulation (VC-SVPWM)-based two-level inverter-fed drives.

Robust Formulations for Handling Uncertainty in Kernel Matrices

We study the problem of uncertainty in the entries of the Kernel matrix, arising in SVM formulation. Using Chance Constraint Programming and a novel large deviation inequality we derive a formulation which is robust to such noise. The resulting formulation applies when the noise is Gaussian, or has finite support. The formulation in general is non-convex, but in several cases of interest it reduces to a convex program. The problem of uncertainty in kernel matrix is motivated from the real world problem of classifying proteins when the structures are provided with some uncertainty. The formulation derived here naturally incorporates such uncertainty in a principled manner leading to significant improvements over the state of the art. 1.

Video coding mode decision as a classification problem

In this paper, we show that it is possible to reduce the complexity of Intra MB coding in H.264/AVC based on a novel chance constrained classifier. Using the pairs of simple mean-variances values, our technique is able to reduce the complexity of Intra MB coding process with a negligible loss in PSNR. We present an alternate approach to address the classification problem which is equivalent to machine learning. Implementation results show that the proposed method reduces encoding time to about 20% of the reference implementation with average loss of 0.05 dB in PSNR.

Use of climate information for farm-level decision making: rainfed groundnut in southern India

The production of rainfed crops in semi-arid tropics exhibits large variation in response to the variation in seasonal rainfall. There are several farm-level decisions such as the choice of cropping pattern, whether to invest in fertilizers, pesticides etc., the choice of the period for planting, plant population density etc. for which the appropriate choice (associated with maximum production or minimum risk) depends upon the nature of the rainfall variability or the prediction for a specific year. In this paper, we have addressed the problem of identifying the appropriate strategies for cultivation of rainfed groundnut in the Anantapur region in a semi-arid part of the Indian peninsula. The approach developed involves participatory research with active collaboration with farmers, so that the problems with perceived need are addressed with the modern tools and data sets available. Given the large spatial variation of climate and soil, the appropriate strategies are necessarily location specific. With the approach adopted, it is possible to tap the detailed location specific knowledge of the complex rainfed ecosystem and gain an insight into the variety of options of land use and management practices available to each category of stakeholders. We believe such a participatory approach is essential for identifying strategies that have a favourable cost-benefit ratio over the region considered and hence are associated with a high chance of acceptance by the stakeholders. (C) 2002 Elsevier Science Ltd. All rights reserved.

Fast Robust Pattern Classification Algorithms for Real Time Neuro-Motor Prosthetic Applications

In this paper, we give a brief review of pattern classification algorithms based on discriminant analysis. We then apply these algorithms to classify movement direction based on multivariate local field potentials recorded from a microelectrode array in the primary motor cortex of a monkey performing a reaching task. We obtain prediction accuracies between 55% and 90% using different methods which are significantly above the chance level of 12.5%.

Molecular Identification and Phylogeny of Bactrocera Species (Diptera: Tephritidae)

Fruit flies that belong to the genus Bactrocera (Diptera: Tephritidae) are major invasive pests of agricultural crops in Asia and Australia. Increased transboundary movement of agricultural produce has resulted in the chance introduction of many invasive species that include Bactrocera mainly as immature stages. Therefore quick and accurate species diagnosis is important at the port of entry, where morphological identification has a limited role, as it requires the presence of adult specimens and the availability of a specialist. Unfortunately when only immature stages are present, a lacunae in their taxonomy impedes accurate species diagnosis. At this juncture, molecular species diagnostics based on COX-I have become handy, because diagnosis is not limited by developmental stages. Yet another method of quick and accurate species diagnosis for Bactrocera spp. is based on the development of species-specific markers. This study evaluated the utility of COX-I for the quick and accurate species diagnosis of eggs, larvae, pupae and adults of B. zonata Saunders, B. tau Walker, and B. dorsalis Hendel. Furthermore the utility of species-specific markers in differentiating B. zonata (500bp) and B. tau (220bp) was shown. Phylogenetic relationships among five subgenera, viz., Austrodacus, Bactrocera, Daculus, Notodacus and Zeugodacus have been resolved employing the 5' region of COX-I (1490-2198); where COX-I sequences for B. dorsalis Hendel, B. tau Walker, B. correcta Bezzi and B. zonata Saunders from India were compared with other NCBI-GenBank accessions. Phylogenetic analysis employing Maximum Parsimony (MP) and Bayesian phylogenetic approach (BP) showed that the subgenus Bactrocera is monophyletic.

Positive Relationships between Association Strength and Phenotypic Similarity Characterize the Assembly of Mixed-Species Bird Flocks Worldwide

Competition theory predicts that local communities should consist of species that are more dissimilar than expected by chance. We find a strikingly different pattern in a multicontinent data set (55 presence-absence matrices from 24 locations) on the composition of mixed-species bird flocks, which are important sub-units of local bird communities the world over. By using null models and randomization tests followed by meta-analysis, we find the association strengths of species in flocks to be strongly related to similarity in body size and foraging behavior and higher for congeneric compared with noncongeneric species pairs. Given the local spatial scales of our individual analyses, differences in the habitat preferences of species are unlikely to have caused these association patterns; the patterns observed are most likely the outcome of species interactions. Extending group-living and social-information-use theory to a heterospecific context, we discuss potential behavioral mechanisms that lead to positive interactions among similar species in flocks, as well as ways in which competition costs are reduced. Our findings highlight the need to consider positive interactions along with competition when seeking to explain community assembly.

Robust formulations for clustering-based large-scale classification

Chebyshev-inequality-based convex relaxations of Chance-Constrained Programs (CCPs) are shown to be useful for learning classifiers on massive datasets. In particular, an algorithm that integrates efficient clustering procedures and CCP approaches for computing classifiers on large datasets is proposed. The key idea is to identify high density regions or clusters from individual class conditional densities and then use a CCP formulation to learn a classifier on the clusters. The CCP formulation ensures that most of the data points in a cluster are correctly classified by employing a Chebyshev-inequality-based convex relaxation. This relaxation is heavily dependent on the second-order statistics. However, this formulation and in general such relaxations that depend on the second-order moments are susceptible to moment estimation errors. One of the contributions of the paper is to propose several formulations that are robust to such errors. In particular a generic way of making such formulations robust to moment estimation errors is illustrated using two novel confidence sets. An important contribution is to show that when either of the confidence sets is employed, for the special case of a spherical normal distribution of clusters, the robust variant of the formulation can be posed as a second-order cone program. Empirical results show that the robust formulations achieve accuracies comparable to that with true moments, even when moment estimates are erroneous. Results also illustrate the benefits of employing the proposed methodology for robust classification of large-scale datasets.

Wet and dry spell characteristics of global tropical rainfall

In this study, we analyze satellite-based daily rainfall observations to compare and contrast the wet and dry spell characteristics of tropical rainfall. Defining a wet (dry) spell as the number of consecutive rainy (nonrainy) days, we find that the distributions of wet spells appear to exhibit universality in the following sense. While both ocean and land regions with high seasonal rainfall accumulation (humid regions; e. g., India, Amazon, Pacific Ocean) show a predominance of 2-4 day wet spells, those regions with low seasonal rainfall accumulation (arid regions; e. g., South Atlantic, South Australia) exhibit a wet spell duration distribution that is essentially exponential in nature, with a peak at 1 day. The behavior that we observed for wet spells is reversed for the dry spell characteristics. In other words, the main contribution to the dry part of the season, in terms of the number of nonrainy days, appears to come from 3-4 day dry spells in the arid regions, as opposed to 1 day dry spells in the humid regions. The total rainfall accumulated in each wet spell has also been analyzed, and we find that the major contribution to seasonal rainfall for arid regions comes from 1-5 day wet spells; however, for humid regions, this contribution comes from wet spells of duration as long as 30 days. We also explore the role of chance as well as the influence of organized convection in determining some of the observed features.