A machine learning assisted optical multistage interconnection network: Performance analysis and hardware demonstration

Sangeetha Rengachary Gopalan,Hemanth Chandran,Nithin Vijayan,Vikas Yadav,Shivam Mishra Electronics and Telecommunications Research Instit 2023 ETRI Journal Vol.45 No.1

Integration of the machine learning (ML) technique in all-optical networks can enhance the effectiveness of resource utilization, quality of service assurances, and scalability in optical networks. All-optical multistage interconnection networks (MINs) are implicitly designed to withstand the increasing highvolume traffic demands at data centers. However, the contention resolution mechanism in MINs becomes a bottleneck in handling such data traffic. In this paper, a select list of ML algorithms replaces the traditional electronic signal processing methods used to resolve contention in MIN. The suitability of these algorithms in improving the performance of the entire network is assessed in terms of injection rate, average latency, and latency distribution. Our findings showed that the ML module is recommended for improving the performance of the network. The improved performance and traffic grooming capabilities of the module are also validated by using a hardware testbed.

Strain induced optical properties of exciton in a CdTe/ZnTe quantum dot

Sangeetha, R.,John Peter, A.,Lee, C.W. Academic Press 2014 Superlattices and microstructures Vol.66 No.-

The influence of strain on the binding energies of heavy and light hole is obtained in a Zn<SUB>x</SUB>Cd<SUB>1-x</SUB>Te/ZnTe quantum dot taking into account the phonon confinement effect. The band offsets are calculated using model-solid approach. The dielectric mismatch effect is introduced between the dielectric constants of dot and the barrier. The strain induced energies of excitonic transitions in a CdZnTe quantum dot with ZnTe barrier are brought out considering the internal electric field induced by the spontaneous and piezoelectric polarizations. Calculations have been obtained using Bessel function as an orthonormal basis for different confinement potentials of barrier height, strain induced linear and third-order nonlinear optical absorption coefficients and the changes of refractive index with the incident photon energy are observed. Our results show that the exciton binding energy is enhanced with the inclusion of potential taking into account the effects of dielectric mismatch and the geometry of quantum dot with various Zinc alloy content has a great influence on the optical properties of the dot.

Effects of strain on the band alignment and the optical gain of a CdTe/ZnTe quantum dot

Sangeetha, R.,Peter, A. John,Yoo, Chang Kyoo Canadian Science Publishing 2014 Canadian journal of physics Vol.92 No.5

<P> The band lineups at the strained layer interfaces, in ZnxCd1-xTe/ZnTe quantum dot nanostructure for various Zn alloy content, are computed using model solid theory with an example of CdTe/ZnTe interface. The effects of strain, due to hydrostatic and biaxial strain, and the internal electric fields, due to the spontaneous and piezoelectric polarization, are taken into consideration. The dielectric mismatch, through the effective potential, is introduced between the dielectric constants of the dot and the barrier materials. The interband emission energy as a function of dot radius is computed for various Zn alloy content. The optical gain spectra of heavy hole exciton for various Zn concentration are studied. Calculations are obtained for different confinement potentials with the inclusion strain effect. Our results show that (i) the potential taking into account the effects of PB potential due to the dielectric mismatch enhances the exciton binding energy and (ii) the geometry of quantum dot, the strain effects and the Zinc alloy content have great influences on the electrical and optical properties of the dot. Our results are in good agreement with the previous investigators. </P>

Does hotel attributes impact customer satisfaction: A sentiment analysis of online reviews

Sangeetha Gunasekar,Sooriya Sudhakar 한국마케팅과학회 2019 마케팅과학연구 Vol.29 No.2

User-generated content is a major source of information particularly in tourism industry where consumers seek unbiased and unregulated information. While making their hotel booking decisions, consumers refer to the previous guests’ experiences expressed in the hotel reviews across social media. Studies in the literature have focused on enhancing the understanding of what makes customers satisfied or dissatisfied. They have analyzed the text reviews and the patterns in the overall rating and ranking of the hotels given by hotel guests. While most emerging destinations have been studied in the literature, studies related to India, a fast-growing leisure destination are scant. The present study tries to understand the hotel attributes that contribute towards customer satisfaction or dissatisfaction using online reviews for all hotels of Andaman & Nicobar Islands in India. Among the eight attributes identified from the most frequently used words in the text reviews, the study finds that while the location has a significant probability of increasing the ratings of both highand low-rated reviews, rooms seem to have the most significant impact on lowering the probability of high scores irrespective of positive or negative sentiment review. The study also finds that guests of luxury hotels rate the hotels significantly higher than the guests of midrange and budget hotels.

Experimental investigation in the selection of blank material during deep drawing process using finite element analysis

Sangeetha N,Brathikan V M,Nitheeshwar R K,Jayabalu S 한양대학교 세라믹연구소 2022 Journal of Ceramic Processing Research Vol.23 No.4

Deep Drawing is the forming process where the shape of the punch is projected to the blank by the application of force withthe help of a die and blank holder. Obtaining a quality product in the deep drawing process mainly depends on thecharacteristics of blank materials. The deep drawing process influences the process parameters like blank holder force,coefficient of friction for the lubricants, etc. The deep drawing process was conducted using a hydraulic press machine witha maximum capacity of 50 Newton for blank materials like cold roll steel, stainless steel, aluminum alloys, and tungsten. Theblank force required for the deep drawing process was investigated on various materials and analysed with the materialproperties. The Finite Element method is used to check the quality of the deep drawing product by applying various forcesobtained in the experimental methods. Shear Stress and Strain obtained in the product during the deep drawing process werestudied and verified with the analytical stress and strain. It was found that ductile material yields a very good drawing ratiowhereas brittle material such as Titanium has very little drawing ratio which could be further enhanced by the combinationof reinforcement or a mixture of allied metals.

Pathological Status of Pyricularia angulata Causing Blast and Pitting Disease of Banana in Eastern India

Sangeetha Ganesan,Hari Shankar Singh,Srinivas Petikam,Debasish Biswal 한국식물병리학회 2017 Plant Pathology Journal Vol.33 No.1

Incidence of leaf blast on nursery plants and pittingdisease on maturing banana bunches has beenrecorded in banana plantations during rainy seasonin Eastern India during 2014 to 2015. Taxonomicalidentification as well as DNA sequence analysis of theinternal transcribed spacer region of fungus isolatedfrom affected tissue culture derived plantlets andfruits confirmed the pathogen to be Pyricularia angulataHashioka “in both the cases”. Koch’s postulateswere proved on young plantlets as well as on maturingfruits of cv. Grand Naine under simulated conditions. Evolutionary history was inferred and presented forour P. angulata strain PG9001 with GenBank accessionno. KU984740. The analysis indicated that theP. angulata is phylogenitically distinct from other relatedspecies related to both Pyricularia and Magnaporthe. Detailed symptoms of blast lesions on youngleaves, transition leaves, mid rib, petioles, peduncle,maturing bunches, bunch stalks and cushions weredocumented. Notably, the distinct small pitting spotson maturing bunches reduced the visual appeal of maturefruits. Appearance of pitting symptoms on fruitsin relation with age of fruits and their distributionpattern on bunch and fingers was also documentedin detail. Further, the roles of transitory leaves, weedhosts, seasonality on disease occurrence have also beendocumented.

Copy Number Variation Burden on Asthma Subgenome in Normal Cohorts Identifies Susceptibility Markers

Sangeetha Vishweswaraiah,Avinash M Veerappa,Padukudru A Mahesh,Sareh R Jahromi,Nallur B. Ramachandra 대한천식알레르기학회 2015 Allergy, Asthma & Immunology Research Vol.7 No.3

Purpose: Asthma is a complex disease caused by interplay of genes and environment on the genome of an individual. Copy number variations (CNVs) are more common compared to the other variations that disrupt genome organization. The effect of CNVs on asthma subgenome has been less studied compared to studies on the other variations. We report the assessments of CNV burden in asthma genes of normal cohorts carried out in different geographical areas of the world and discuss the relevance of the observation with respect to asthma pathogenesis. Methods: CNV analysis was performed using Affymerix high-resolution arrays, and various bioinformatics tools were used to understand the influence of genes on asthma pathogenesis. Results: This study identified 61 genes associated with asthma and provided various mechanisms and pathways underlying asthma pathogenesis. CCL3L1, ADAM8, and MUC5B were the most prevalent asthma genes. Among them, CCL3L1 was found across all 12 populations in varying copy number states. This study also identified the inheritance of asthma-CNVs from parents to offspring creating the latent period for manifestation of asthma. Conclusions: This study revealed CNV burden with varying copy number states and identified susceptibility towards the disease manifestation. It can be hypothesized that primary CNVs may not be the initiating event in the pathogenesis of asthma and additional preceding mutations or CNVs may be required. The initiator or primary CNVs sensitize normal cohorts leading to an increased probability of accumulating mutations or exposure to allergic stimulating agents that can augment the development of asthma.

Statistical Characterization of the Multi-Charged Fragment Ions in the CID and HCD Spectrum

( Sangeetha Ramachandran ),( Tessamma Thomas ) 한국질량분석학회 2021 Mass spectrometry letters Vol.12 No.2

Collision-induced dissociation (CID) and higher-energy collisional dissociation (HCD) are the widely used fragmentation technique in mass spectrometry-based proteomics studies. Understanding the fragmentation pattern from the tandem mass spectra using statistical methods helps to implement efficient spectrum analysis algorithms. The study characterizes the frequency of occurrence of multi-charged fragment ions and their neutral loss events of doubly and triply charged peptides in the CID and HCD spectrum. The dependency of the length of the fragment ion on the occurrence of multi-charged fragment ion is characterized here. Study shows that the singly charged fragment ions are generally dominated in the doubly charged peptide spectrum. However, as the length of the product ion increases, the frequency of occurrence of charge 2 fragment ions increases. The y- ions have more tendencies to generate charge 2 fragment ions than b- ions, both in CID and HCD spectrum. The frequency of occurrence of charge 2 fragment ion peaks is prominent upon the dissociation of the triply charged peptides. For triply charged peptides, product ion of higher length occurred in multiple charge states in CID spectrum. The neutral loss peaks mostly exist in charge 2 states in the triply charged peptide spectrum. The b-ions peaks are observed in much less frequency than y-ions in HCD spectrum as the length of the fragment increases. Isotopic peaks are occurred in charge 2 state both in doubly and triply charged peptide’s HCD spectrum.

A Novel Accurate and Time Efficient Map Reduce Approach for Biomedical Ontology Alignment

Sangeetha Balachandran,Vidhyapriya Ranganathan 대한전기학회 2024 Journal of Electrical Engineering & Technology Vol.19 No.3

This research focuses on enhancing ontology mapping in the biomedical domain by reducing execution time and automating the mapping process. Biomedical ontologies are crucial for accurate interpretation of medical records and decision making. However, two main challenges are overlapping concepts and a large search space. To address these issues, a distributed environment using the Hadoop framework is implemented, and a Map-Reduce algorithm is employed for parallelizing the mapping system, resulting in signifcant time reduction. For precise alignment, an Extreme Learning Machine based neural network is utilized. Evaluation is conducted using OAEI and OBO biomedical ontologies, with simulation results demonstrating notable improvements in execution time and evaluation metrics using the proposed ontology mapping system's multi-strategy similarity metrics. The mapping between ontologies is represented using the Alignment API, simplifying the calculation process.

On f-cosymplectic and (k, µ)-cosymplectic Manifolds Admitting Fischer-Marsden Conjecture

Sangeetha Mahadevappa,Halammanavar Gangadharappa Nagaraja 경북대학교 자연과학대학 수학과 2023 Kyungpook mathematical journal Vol.63 No.3

The aim of this paper is to study the Fisher-Marsden conjucture on f cosymplectic and (k, µ)-cosymplectic manifolds. First we prove that a compact f cosymplectic manifold satisfying the Fisher-Marsden equation R′*g = 0 is either an Einstein manifold or a local product of a Kahler manifold and an interval or a unit circle S1 . Further we obtain that in almost (k, µ)-cosymplectic manifolds with k < 0, the Fisher-Marsden equation has a trivial solution.