Semi-supervised SVMs for classification with unknown class proportions and a small labeled dataset

In the design of practical web page classification systems one often encounters a situation in which the labeled training set is created by choosing some examples from each class; but, the class proportions in this set are not the same as those in the test distribution to which the classifier will be actually applied. The problem is made worse when the amount of training data is also small. In this paper we explore and adapt binary SVM methods that make use of unlabeled data from the test distribution, viz., Transductive SVMs (TSVMs) and expectation regularization/constraint (ER/EC) methods to deal with this situation. We empirically show that when the labeled training data is small, TSVM designed using the class ratio tuned by minimizing the loss on the labeled set yields the best performance; its performance is good even when the deviation between the class ratios of the labeled training set and the test set is quite large. When the labeled training data is sufficiently large, an unsupervised Gaussian mixture model can be used to get a very good estimate of the class ratio in the test set; also, when this estimate is used, both TSVM and EC/ER give their best possible performance, with TSVM coming out superior. The ideas in the paper can be easily extended to multi-class SVMs and MaxEnt models.

Assessment of hydrologic impacts of climate change in Tunga-Bhadra river basin, India with HEC-HMS and SDSM

Climate change would significantly affect many hydrologic systems, which in turn would affect the water availability, runoff, and the flow in rivers. This study evaluates the impacts of possible future climate change scenarios on the hydrology of the catchment area of the TungaBhadra River, upstream of the Tungabhadra dam. The Hydrologic Engineering Center's Hydrologic Modeling System version 3.4 (HEC-HMS 3.4) is used for the hydrological modelling of the study area. Linear-regression-based Statistical DownScaling Model version 4.2 (SDSM 4.2) is used to downscale the daily maximum and minimum temperature, and daily precipitation in the four sub-basins of the study area. The large-scale climate variables for the A2 and B2 scenarios obtained from the Hadley Centre Coupled Model version 3 are used. After model calibration and testing of the downscaling procedure, the hydrological model is run for the three future periods: 20112040, 20412070, and 20712099. The impacts of climate change on the basin hydrology are assessed by comparing the present and future streamflow and the evapotranspiration estimates. Results of the water balance study suggest increasing precipitation and runoff and decreasing actual evapotranspiration losses over the sub-basins in the study area.

The influence of debris cover and glacial lakes on the recession of glaciers in Sikkim Himalaya, India

We investigated area changes in glaciers covering an area of similar to 200 km(2) in the Tista basin, Sikkim, Eastern Indian Himalaya, between similar to 1990 and 2010 using Landsat Thematic Mapper (TM) and Indian Remote-sensing Satellite (IRS) images and related the changes to debris cover, supraglacial lakes and moraine-dam lakes. The glaciers lost an area of 3.3 +/- 0.8% between 1989/90 and 2010. More detailed analysis revealed an area loss of 2.00 +/- 0.82, 2.56 +/- 0.61 and 2.28 +/- 2.01 km(2) for the periods 1989-97, 1997-2004/05 and 2004-2009/10, respectively. This indicates an accelerated retreat of glaciers after 1997. On further analysis, we observed (1) the formation and expansion of supraglacial lakes on many debris-covered glaciers and (2) the merging of these lakes over time, leading to the development of large moraine-dam lakes. We also observed that debris-covered glaciers with lakes lose a greater area than debris-covered glaciers without lakes and debris-free glaciers. The climatic data for 24 years (1987-2011), measured at the Gangtok meteorological station (1812 m a.s.l.), showed that the region experienced a 1.0 degrees C rise in the summer minimum temperature and a 2.0 degrees C rise in the winter minimum temperature, indicating hotter summers and warmer winters. There was no significant trend in the total annual precipitation. We find that glacier retreat is caused mainly by a temperature increase and that debris-covered glaciers can retreat at a faster rate than debris-free glaciers, if associated with lakes.

In vitro and in silico biological studies of novel thiazolo3,2-a]pyrimidine-6-carboxylate derivatives

In the present study, we report the synthesis, characterization of new series of thiazolo3,2-a]pyrimidine-6-carboxylate derivatives 3a-f and 4a-f. The newly synthesized compounds were screened for in vitro antimicrobial and antiviral activities. The probable mode of action of these active compounds was determined through in silico docking study by docking the receptor methionyl-tRNA synthetase and human inosine-5'-monophosphate dehydrogenase (IMPDH) for antibacterial and antiviral activities, respectively. Among the compounds, 4c exhibited excellent in vitro antimicrobial activity against all tested strains with binding and docking energies -35.6 and -12.4 kcal/mol, respectively. The antiviral studies were carried out for the selected compounds in which 4a exhibited 73.69 and 54.42 % of inhibition of buffalopox and camelpox viruses, respectively. Furthermore, compound 4a showed minimum docking and binding energy along with the maximum hydrogen/hydrophobic interaction with IMPDH. The study contributes towards identification and screening of potential antimicrobial and antiviral agent's against the pathogens.

Layered Exact-Repair Regenerating Codes via Embedded Error Correction and Block Designs

A new class of exact-repair regenerating codes is constructed by stitching together shorter erasure correction codes, where the stitching pattern can be viewed as block designs. The proposed codes have the help-by-transfer property where the helper nodes simply transfer part of the stored data directly, without performing any computation. This embedded error correction structure makes the decoding process straightforward, and in some cases the complexity is very low. We show that this construction is able to achieve performance better than space-sharing between the minimum storage regenerating codes and the minimum repair-bandwidth regenerating codes, and it is the first class of codes to achieve this performance. In fact, it is shown that the proposed construction can achieve a nontrivial point on the optimal functional-repair tradeoff, and it is asymptotically optimal at high rate, i.e., it asymptotically approaches the minimum storage and the minimum repair-bandwidth simultaneously.

Ischaemic concentrations of lactate increase TREK1 channel activity by interacting with a single histidine residue in the carboxy terminal domain

Key points The physiological metabolite, lactate and the two-pore domain leak potassium channel, TREK1 are known neuroprotectants against cerebral ischaemia. However, it is not known whether lactate interacts with TREK1 channel to provide neuroprotection. In this study we show that lactate increases TREK1 channel activity and hyperpolarizes CA1 stratum radiatum astrocytes in hippocampal slices. Lactate increases open probability and decreases longer close time of the human (h)TREK1 channel in a concentration dependent manner. Lactate interacts with histidine 328 (H328) in the carboxy terminal domain of hTREK1 channel to decrease its dwell time in the longer closed state. This interaction was dependent on the charge on H328. Lactate-insensitive mutant H328A hTREK1 showed pH sensitivity similar to wild-type hTREK1, indicating that the effect of lactate on hTREK1 is independent of pH change. AbstractA rise in lactate concentration and the leak potassium channel TREK1 have been independently associated with cerebral ischaemia. Recent literature suggests lactate to be neuroprotective and TREK1 knockout mice show an increased sensitivity to brain and spinal cord ischaemia; however, the connecting link between the two is missing. Therefore we hypothesized that lactate might interact with TREK1 channels. In the present study, we show that lactate at ischaemic concentrations (15-30mm) at pH7.4 increases TREK1 current in CA1 stratum radiatum astrocytes and causes membrane hyperpolarization. We confirm the intracellular action of lactate on TREK1 in hippocampal slices using monocarboxylate transporter blockers and at single channel level in cell-free inside-out membrane patches. The intracellular effect of lactate on TREK1 is specific since other monocarboxylates such as pyruvate and acetate at pH7.4 failed to increase TREK1 current. Deletion and point mutation experiments suggest that lactate decreases the longer close dwell time incrementally with increase in lactate concentration by interacting with the histidine residue at position 328 (H328) in the carboxy terminal domain of the TREK1 channel. The interaction of lactate with H328 is dependent on the charge on the histidine residue since isosteric mutation of H328 to glutamine did not show an increase in TREK1 channel activity with lactate. This is the first demonstration of a direct effect of lactate on ion channel activity. The action of lactate on the TREK1 channel signifies a separate neuroprotective mechanism in ischaemia since it was found to be independent of the effect of acidic pH on channel activity. Key points The physiological metabolite, lactate and the two-pore domain leak potassium channel, TREK1 are known neuroprotectants against cerebral ischaemia. However, it is not known whether lactate interacts with TREK1 channel to provide neuroprotection. In this study we show that lactate increases TREK1 channel activity and hyperpolarizes CA1 stratum radiatum astrocytes in hippocampal slices. Lactate increases open probability and decreases longer close time of the human (h)TREK1 channel in a concentration dependent manner. Lactate interacts with histidine 328 (H328) in the carboxy terminal domain of hTREK1 channel to decrease its dwell time in the longer closed state. This interaction was dependent on the charge on H328. Lactate-insensitive mutant H328A hTREK1 showed pH sensitivity similar to wild-type hTREK1, indicating that the effect of lactate on hTREK1 is independent of pH change.