Search

This study aims to examine how Sub-Saharan Africa’s top ranked companies themselves convey their role as social-economic development partners and/or agents in their corporate communications, and to what extent country- and industry-specific characteristics influence their reporting, i.e., via CSR websites reporting. Using Forbes’ 2,000 world-largest-corporations ranking, based on 2017 ranking list, we select the largest corporations. We then filter our search to select Africa-largest-corporations. A content analysis of the CSR reports and/or Websites of the sample companies was conducted to identify the motives for CSR practices, CSR managerial processes and stakeholder issues addressed in CSR reporting. Several of the “motives underlying CSR practices”, “CSR managerial processes” and “stakeholder issues addressed” in CSR reporting appear to converge around similar themes (or issues), given that all three sample companies operate in the financial services industry within the specificities of South Africa’s context. The findings suggest that the specificities of South Africa’s historical development and/or socio-cultural realities, legislation and industry charters are important factors that influence CSR (or sustainability) practices in the context of Sub-Saharan Africa. Although, studies that mainly focus on the contents of corporate website are uncommon, this study has limitations as we solely relied on publicly accessible CSR reporting, rather than, for example, interviewing employees, customers (or clients) and/or regulators to verify the claims by the sample companies. As this study examines CSR website reporting practices by organisations, it provides a useful insight for competitor benchmarking that can be used by organisations to imrpove their CSR website reporting practices. This study contributes to extant CSR research as its provides empirical evidence of the contextuality of CSR in Sub-Saharan Africa, as well as explores how country- and industry-spec

Large Language Models (LLMs) have demonstrated great potential for code generation and other software engineering tasks. Vulnerability detection is of crucial importance to maintaining the security, integrity, and trustworthiness of software systems. Precise vulnerability detection requires reasoning about the code, making it a good case study for exploring the limits of LLMs' reasoning capabilities. Although recent work has applied LLMs to vulnerability detection using generic prompting techniques, their full capabilities for this task and the types of errors they make when explaining identified vulnerabilities remain unclear. In this paper, we surveyed eleven LLMs that are state-of-the-art in code generation and commonly used as coding assistants, and evaluated their capabilities for vulnerability detection. We systematically searched for the best-performing prompts, incorporating techniques such as in-context learning and chain-of-thought, and proposed three of our own prompting methods. Our results show that while our prompting methods improved the models' performance, LLMs generally struggled with vulnerability detection. They reported 0.5-0.63 Balanced Accuracy and failed to distinguish between buggy and fixed versions of programs in 76% of cases on average. By comprehensively analyzing and categorizing 287 instances of model reasoning, we found that 57% of LLM responses contained errors, and the models frequently predicted incorrect locations of buggy code and misidentified bug types. LLMs only correctly localized 6 out of 27 bugs in DbgBench, and these 6 bugs were predicted correctly by 70-100% of human participants. These findings suggest that despite their potential for other tasks, LLMs may fail to properly comprehend critical code structures and security-related concepts. Our data and code are available at https://figshare.com/s/78fe02e56e09ec49300b.

Recent breakthroughs in artificial intelligence offer tremendous promise for the development of self-driving applications. Deep Neural Networks, in particular, are being utilized to support the operation of semi-autonomous cars through object identification and semantic segmentation. To assess the inadequacy of the current dataset in the context of autonomous and semi-autonomous cars, we created a new dataset named ANNA. This study discusses a custom-built dataset that includes some unidentified vehicles in the perspective of Bangladesh, which are not included in the existing dataset. A dataset validity check was performed by evaluating models using the Intersection Over Union (IOU) metric. The results demonstrated that the model trained on our custom dataset was more precise and efficient than the models trained on the KITTI or COCO dataset concerning Bangladeshi traffic. The research presented in this paper also emphasizes the importance of developing accurate and efficient object detection algorithms for the advancement of autonomous vehicles.

We provide the identification and species delineation of this biocontrol agent as Stomphastis thraustica (Meyrick in Trans Ent Soc Lond 80(1):107–120, 1908) belonging to the family Gracillariidae. We clarify the distribution pattern of S. thraustica, its host plant preferences, and present taxonomic and molecular diagnoses based on original morphological and genetic data as well as data retrieved from historic literature and genetic databases. Following our own collecting efforts in three continents Africa, South America, and Australia as well as our study of historic museum collection material, we present many new distribution records of S. thraustica for countries and territories in the world including the new discovery of this species in the Neotropical region and we report its introduction in Australia as a biocontrol agent. Using mitogenomic and COI gene data, we clarified that the closest relative of S. thraustica is Stomphastis sp. that occurs in Madagascar and Australia and feeds on the same host plant as S. thraustica – Jatropha gossypiifolia L. (Euphorbiaceae). The molecular sequence divergence in the mitochondrial DNA barcode fragment between these two closely related species S. thraustica and Stomphastis sp. is over 5.7% supporting that they are different species.

The impact assessment of a weed biocontrol agent offers insight into the potential of the agent and weed responses in long-term weed management. The impact of the jewel beetle Hedwigiella jureceki Obenberger (Coleoptera: Buprestidae) and its first-generation larvae on the long-pod (LP) and short-pod (SP) forms of cat's claw creeper, Dolichandra unguis-cati (L.) L.G. Lohmann (Bignoniaceae) was studied by initiating one, two, or five pairs (male and female) of adults in a glasshouse experiment. With the higher density of adults, the number of leaves increased significantly, as did the numbers of dead leaves and larval mines, and the percentages of affected leaves (up to 40%), and damaged leaf areas (up to 60%). Plants with the five pairs of adults had fewer leaves and reduced dry weights (e.g., tuber biomass was reduced by 50.4–62.5%). Regardless of adult density, the dry weights of the leaves and tubers were lower in the SP form. Conversely, the LP form had fewer leaves (ca. 17% fewer) and slower shoot growth (mean ± SE = 46.3 ± 2.6 cm per week) than the SP form (66.5 ± 3.8 cm per week). However, the number of tubers was not affected by the plant form and adult density. Larval mines (262 ± 30), pupae (186 ± 17), and adults (144 ± 14) in LP form were higher than in SP form (243 ± 20, 167 ± 14, and 115 ± 11, respectively). Likewise, the average weight of adults that emerged on the LP form was higher. Therefore, H. jureceki impacted both LP and SP forms. The SP form showed a higher impact in terms of reduced tuber growth, and the LP form is anticipated to be harmed more in the long term as a higher number of H. jureceki offspring developed on it when competition for food was increased.

Over the last few years, social media has evolved into a medium for expressing personal views, emotions, and even business and political proposals, recommendations, and advertisements. We address the topic of identifying emotions from text data obtained from social media posts like Twitter in this research. We have deployed different traditional machine learning techniques such as Support Vector Machines (SVM), Naive Bayes, Decision Trees, and Random Forest, as well as deep neural network models such as LSTM, CNN, GRU, BiLSTM, BiGRU to classify these tweets into four emotion categories (Fear, Anger, Joy, and Sadness). Furthermore, we have constructed a BiLSTM and BiGRU ensemble model. The evaluation result shows that the deep neural network models(BiGRU, to be specific) produce the most promising results compared to traditional machine learning models, with an 87.53 % accuracy rate. The ensemble model performs even better (87.66 %), albeit the difference is not significant. This result will aid in the development of a decision-making tool that visualizes emotional fluctuations., Comment: Course Project

In recent days, the number of technology enthusiasts is increasing day by day with the prevalence of technological products and easy access to the internet. Similarly, the amount of people working behind this rapid development is rising tremendously. Computer programmers consist of a large portion of those tech-savvy people. Codeforces, an online programming and contest hosting platform used by many competitive programmers worldwide. It is regarded as one of the most standardized platforms for practicing programming problems and participate in programming contests. In this research, we propose a framework that predicts the performance of any particular contestant in the upcoming competitions as well as predicts the rating after that contest based on their practice and the performance of their previous contests., Comment: Camera Ready Version

Sluggish oxygen reduction reaction (ORR) of electrodes is one of the main challenges in fuel cell systems. This study explored the kinetics of the ORR reaction mechanism, which enables us to understand clearly the electrochemical activity of the electrode. In this research, electrocatalysts were synthesized from platinum (Pt) catalyst with multi-walled carbon nanotubes (MWCNTs) coated by three polymers (polybenzimidazole (PBI), sulfonated tetrafluoroethylene (Nafion), and polytetrafluoroethylene (PTFE)) as the supporting materials by the polyol method while hexachloroplatinic acid (H2PtCl6) was used as a catalyst precursor. The oxygen reduction current of the synthesized electrocatalysts increased that endorsed by linear sweep voltammetry (LSV) curves while increasing the rotation rates of the disk electrode. Additionally, MWCNT-PBI-Pt was attributed to the maximum oxygen reduction current densities at −1.45 mA/cm2 while the minimum oxygen reduction current densities of MWCNT-Pt were obtained at −0.96 mAcm2. However, the ring current densities increased steadily from potential 0.6 V to 0.0 V due to their encounter with the hydrogen peroxide species generated by the oxygen reduction reactions. The kinetic limiting current densities (JK) increased gradually with the applied potential from 1.0 V to 0.0 V. It recommends that the ORR consists of a single step that refers to the first-order reaction. In addition, modified MWCNT-supported Pt electrocatalysts exhibited high electrochemically active surface areas (ECSA) at 24.31 m2/g of MWCNT-PBI-Pt, 22.48 m2/g of MWCNT-Nafion-Pt, and 20.85 m2/g of MWCNT-PTFE-Pt, compared to pristine MWCNT-Pt (17.66 m2/g). Therefore, it can be concluded that the additional ionomer phase conducting the ionic species to oxygen reduction in the catalyst layer could be favorable for the ORR reaction.

Recently, pretrained language models have shown state-of-the-art performance on the vulnerability detection task. These models are pretrained on a large corpus of source code, then fine-tuned on a smaller supervised vulnerability dataset. Due to the different training objectives and the performance of the models, it is interesting to consider whether the models have learned the semantics of code relevant to vulnerability detection, namely bug semantics, and if so, how the alignment to bug semantics relates to model performance. In this paper, we analyze the models using three distinct methods: interpretability tools, attention analysis, and interaction matrix analysis. We compare the models' influential feature sets with the bug semantic features which define the causes of bugs, including buggy paths and Potentially Vulnerable Statements (PVS). We find that (1) better-performing models also aligned better with PVS, (2) the models failed to align strongly to PVS, and (3) the models failed to align at all to buggy paths. Based on our analysis, we developed two annotation methods which highlight the bug semantics inside the model's inputs. We evaluated our approach on four distinct transformer models and four vulnerability datasets and found that our annotations improved the models' performance in the majority of settings - 11 out of 16, with up to 9.57 points improvement in F1 score compared to conventional fine-tuning. We further found that with our annotations, the models aligned up to 232% better to potentially vulnerable statements. Our findings indicate that it is helpful to provide the model with information of the bug semantics, that the model can attend to it, and motivate future work in learning more complex path-based bug semantics. Our code and data are available at https://figshare.com/s/4a16a528d6874aad51a0.

Deep learning vulnerability detection has shown promising results in recent years. However, an important challenge that still blocks it from being very useful in practice is that the model is not robust under perturbation and it cannot generalize well over the out-of-distribution (OOD) data, e.g., applying a trained model to unseen projects in real world. We hypothesize that this is because the model learned non-robust features, e.g., variable names, that have spurious correlations with labels. When the perturbed and OOD datasets no longer have the same spurious features, the model prediction fails. To address the challenge, in this paper, we introduced causality into deep learning vulnerability detection. Our approach CausalVul consists of two phases. First, we designed novel perturbations to discover spurious features that the model may use to make predictions. Second, we applied the causal learning algorithms, specifically, do-calculus, on top of existing deep learning models to systematically remove the use of spurious features and thus promote causal based prediction. Our results show that CausalVul consistently improved the model accuracy, robustness and OOD performance for all the state-of-the-art models and datasets we experimented. To the best of our knowledge, this is the first work that introduces do calculus based causal learning to software engineering models and shows it's indeed useful for improving the model accuracy, robustness and generalization. Our replication package is located at https://figshare.com/s/0ffda320dcb96c249ef2., Comment: ICSE 2024, Camera Ready Version

The gall-inducing tephritid fly Notomma mutilum (Bezzi) (Diptera: Tephritidae) from Senegal is a prospective biological control agent of prickly acacia, Vachellia nilotica subsp. indica (Benth.) Kyal. & Boatwr (Fabaceae) in Australia. Wide variation in the oviposition behaviour and life cycle of N. mutilum in quarantine resulted in difficulties with colony maintenance and host specificity testing. To identify the factors responsible, adults that emerged from galls cut from plants 100 days after oviposition (enforced population, EP) were compared with adults that emerged from galls on the plants (normal population, NP). Oviposition by females of different ages did not vary significantly. The EP females had more variable egg-laying (galls on 46.8% plants) and made fewer oviposition scars (≈37 per female) than NP females (galls on >90% plants and ≈95 scars per female). However, the life cycle duration of EP progeny was shorter (140.7?±?1.7 days) and less variable than NP progeny (160.1?±?12.8 days). Egg-laying on all the plants was achieved when a greater number of EP adults (three pairs) were used. Actively growing tender shoots in the upper crown of the older host plants were preferred sites for oviposition. Galls were initiated (3.5?±?0.2 days) before egg hatching (11.3?±?0.3 days), and bigger (longer and wider) galls were observed on the preferred oviposition sites. Therefore, plants with actively growing shoots in the upper crown should be selected for host specificity testing and three pairs of N. mutilum adults should be used.

We provide the identification and species delineation of this biocontrol agent as Stomphastis thraustica (Meyrick in Trans Ent Soc Lond 80(1):107–120, 1908) belonging to the family Gracillariidae. We clarify the distribution pattern of S. thraustica, its host plant preferences, and present taxonomic and molecular diagnoses based on original morphological and genetic data as well as data retrieved from historic literature and genetic databases. Following our own collecting efforts in three continents Africa, South America, and Australia as well as our study of historic museum collection material, we present many new distribution records of S. thraustica for countries and territories in the world including the new discovery of this species in the Neotropical region and we report its introduction in Australia as a biocontrol agent. Using mitogenomic and COI gene data, we clarified that the closest relative of S. thraustica is Stomphastis sp. that occurs in Madagascar and Australia and feeds on the same host plant as S. thraustica – Jatropha gossypiifolia L. (Euphorbiaceae). The molecular sequence divergence in the mitochondrial DNA barcode fragment between these two closely related species S. thraustica and Stomphastis sp. is over 5.7% supporting that they are different species.

The leaf-mining moth, Stomphastis thraustica (Meyrick, 1908) was imported to Australia as a potential biological control agent of an exotic weed, bellyache bush (Jatropha gossypiifolia), from Peru. The insect colony has been maintained in the quarantine facility for over eight years but recently, significant mortality was observed in the culture. The larvae demonstrated swollen intersegments with a fragile integument. The infected larvae are cloudy muted green or yellowish whereas a healthy late instar larva is a vivid green. They slowly dehydrate and eventually die, at which point the larval body becomes rubbery and turns to black. We used next generation sequencing to identify the cause of mortality in the insects. Total RNA was extracted from 20 larvae in two cohorts, one with and one without apparent symptoms of disease, for deep sequencing on NovaSeq platform after eukaryote ribosomal RNA depletion. We identified several non-insect sequences belonging to viruses, bacteria, and fungi, but none of those showed significant abundance or enrichment in the infected dataset. The sequences related to a unicellular yeast, Saccharomyces cerevisiae, and they were among the highly expressed non-insect contigs; more than 5% of reads in both libraries mapped to the genome of this opportunistic microorganism.

The concept of Internet of Things (IoT) has been widely studied in smart home networks, smart city networks, smart grid systems, autonomous driving systems, and smart healthcare systems. In IoT, the IEEE 802.11n wireless local-area network (WLAN) is used as a common communication technology due to its exibility and low cost. Then, the high performance WLAN is required to enhance quality of service (QoS) of large-scale IoT applications connecting a number of devices or sensors allocated in wide areas. WLAN can use the limited number of partially overlapping channels (POCs) at 2.4 GHz band. The WLAN performance can be degraded by interfered signals from other WLANs. Then, to optimize the POC assignment by reducing interferences, we have proposed the throughput drop estimation model for concurrently communicating multiple links under interferences. Unfortunately, the 40 MHz channel bonding (CB) and the 20 MHz non-CB are considered separately, while the transmission power is always fixed to the maximum. In this paper, we study the throughput drop estimation model under coexistence of CB and non-CB while the transmission power is changed. Then, we present its application to the joint optimization of assigning the transmission power, the frequency channel, and the channel bonding to enhance the throughput performance of IEEE 802.11n WLAN. For evaluations, we compare estimated throughputs by the model with measured ones in various network topologies to verify the model accuracy. Then, we apply the model to the joint assignment optimization in them, and confirm the effectiveness through simulations and experiments using the testbed system.

Nowadays, the IEEE 802.11 wireless local area network (WLAN) has been widely used for Internet access services around the world. Then, the unfairness or insufficiency in meeting the throughput request can appear among concurrently communicating hosts with the same access point (AP), which should be solved by sacrificing advantageous hosts. Previously, we studied the fairness control method by adopting packet transmission delay at the AP. However, it suffers from slow convergence and may not satisfy different throughput requests among hosts. In this paper, we propose a throughput request satisfaction method for providing fair or different throughput requests when multiple hosts are concurrently communicating with a single AP. To meet the throughput request, the method (1) measures the single and concurrent throughput for each host, (2) calculates the channel occupying time from them, (3) derives the target throughput to achieve the given throughput request, and (4) controls the traffic by applying traffic shaping at the AP. For evaluations, we implemented the proposal in the WLAN testbed system with one Raspberry Pi AP and up to five hosts, and conducted extensive experiments in five scenarios with different throughput requests. The results confirmed the effectiveness of our proposal.

Deep learning (DL) models of code have recently reported great progress for vulnerability detection. In some cases, DL-based models have outperformed static analysis tools. Although many great models have been proposed, we do not yet have a good understanding of these models. This limits the further advancement of model robustness, debugging, and deployment for the vulnerability detection. In this paper, we surveyed and reproduced 9 state-of-the-art (SOTA) deep learning models on 2 widely used vulnerability detection datasets: Devign and MSR. We investigated 6 research questions in three areas, namely model capabilities, training data, and model interpretation. We experimentally demonstrated the variability between different runs of a model and the low agreement among different models' outputs. We investigated models trained for specific types of vulnerabilities compared to a model that is trained on all the vulnerabilities at once. We explored the types of programs DL may consider "hard" to handle. We investigated the relations of training data sizes and training data composition with model performance. Finally, we studied model interpretations and analyzed important features that the models used to make predictions. We believe that our findings can help better understand model results, provide guidance on preparing training data, and improve the robustness of the models. All of our datasets, code, and results are available at https://doi.org/10.6084/m9.figshare.20791240., Comment: 12 pages, 14 figures. Accepted at ICSE 2023. Camera-ready version

Improved sanitation has been hypothesized to reduce soil-transmitted helminth (STH) infections by reducing the prevalence and concentration of STH eggs/larvae in soil. We evaluated the effect of a randomized sanitation program (providing households with an improved dual-pit latrine, tools for child/animal feces management, and behavioral messaging) on reducing the prevalence and concentration of STH eggs in soil from household courtyards. We collected soil samples from 1405 households enrolled in the sanitation intervention (n = 419) and control (n = 914) groups of a cluster-randomized controlled trial (WASH Benefits) in rural Bangladesh approximately 2 years after the initiation of the interventions. We analyzed samples for Ascaris lumbricoides, Trichuris trichiura, and hookworm eggs by microscopy. We estimated prevalence ratios (PR) and egg count ratio (ECR) to compare the prevalence of STH eggs and arithmetic and geometric mean egg counts for STH eggs per gram of soil in the sanitation and control arms. Among intervention households, latrines achieved high and sustained user uptake by adults while child open defecation remained common and most households did not dispose of child feces hygienically. In courtyard soil from control households, the prevalence of any STH eggs was 75.7% and the prevalence of any larvated STH eggs was 67.3%. A. lumbricoides was detected in 63.0% of control samples and T. trichiura in 55.7% of control samples; hookworm was not detected in any sample. In the control arm, the arithmetic mean egg count for any STH was 3.96 eggs/dry gram, while the geometric mean was 1.58 eggs/dry gram. There was no difference between the intervention and control groups in the prevalence of any STH eggs (PR = 0.98 (95% CI: 0.91, 1.05)) or mean egg counts (ECR = 0.08 (95% CI: -0.10, 0.26) for geometric mean and 0.07 (95% CI: -0.22, 0.37) for arithmetic mean). Adjusted models gave similar results. A compound-level sanitation intervention that provided improved

Solar powered canopy style smart device charging station is a new and very unique idea in the context of Bangladesh. Student or other official personnel of any public or private organization will be beneficial if they find their work place, equipped with such useful, sustainable power station with sitting lounges where they can relax, can do group studies or work and can charge their smart devices i.e.; laptops and smart phones in an environment friendly way using renewable solar energy. With this point of view, this paper represents development of a canopy style solar energy based smart device charging station having capacity of 600W, with 200 Ah battery backup at Patuakhali Science and Technology University campus. This could be an iconic idea of using green energy and to serve people with electrical energy conserving the natural ecosystem harmlessly.

WRINKLED1 (WRI1), a member of the APETALA2 (AP2) class of transcription factors regulates fatty acid biosynthesis and triacylglycerol (TAG) accumulation in plants. Among the four known Arabidopsis WRI1 paralogs, only WRI2 was unable to complement and restore fatty acid content in wri1-1 mutant seeds. Avocado (Persea americana) mesocarp, which accumulates 60-70% dry weight oil content, showed high expression levels for orthologs of WRI2, along with WRI1 and WRI3, during fruit development. While the role of WRI1 as a master regulator of oil biosynthesis is well-established, the function of WRI1 paralogs is poorly understood. Comprehensive and comparative in silico analyses of WRI1 paralogs from avocado (a basal angiosperm) with higher angiosperms Arabidopsis (dicot), maize (monocot) revealed distinct features. Predictive structural analyses of the WRI orthologs from these three species revealed the presence of AP2 domains and other highly conserved features, such as intrinsically disordered regions associated with predicted PEST motifs and phosphorylation sites. Additionally, avocado WRI proteins also contained distinct features that were absent in the nonfunctional Arabidopsis ortholog AtWRI2. Through transient expression assays, we demonstrated that both avocado WRI1 and WRI2 are functional and drive TAG accumulation in Nicotiana benthamiana leaves. We predict that the unique features and activities of ancestral PaWRI2 were likely lost in orthologous genes such as AtWRI2 during evolution and speciation, leading to at least partial loss of function in some higher eudicots. This study provides us with new targets to enhance oil biosynthesis in plants.

Antimicrobial metabolites are produced as secondary metabolites by microorganism as well as the plant growth promoting rhizobacteria. These compounds are widely distributed in nature, where they play an important role in regulating the microbial population of soil, water, sewage and compost. In the present investigation, some antimicrobial metabolites such as antibiotics, siderophores, and HCN were detected in vitro by TLC, CAS agar universal test plates and filter paper with alkaline picrate solution method respectively from the two rhizobacterial strains Pseudomonas aeruginosa UPMP3 and Burkholderia cepacia UPMB3. In vitro bioassay was carried out through antagonistic activity test against G. boninense based on the percentage inhibition of radial growth (PIRG). The strains showed antifungal activity against Ganoderma boninense that is responsible for the disease of basal stem rot in oil palm. The crude extracts obtained from ethyl acetate solvent extraction and analyzed by thin layer chromatography (TLC). Six different antibiotics were detected with different retention factors (Rf) on TLC plates. The Rf values were calculated as 0.88, 0.77, 0.63, 0.53,0.47, 0.28 and 0.23 for 2-4 DAPG, pyoluteorin, phenazine, pyocyanin, phenazine-1-carboxamide (PCN) and pyrrolnitrin successively developed with different solvent systems. Among the different solvent systems ethyl acetate: chloroform was most effective in separating the active bands from the extracts. Siderophores were detected through colour change into blue to orange and HCN was in dark brown to red colour. The antagonistic activity of Pseudomonas aeruginosa UPMP3 and Burkholderia cepacia UPMB3 was evaluated. In case of bacterial antibiotics and volatile and non-volatile effects, the strain P. aeruginosa UPMP3 showed the highest 94.21% and 51.16% inhibitory on the mycelial growth of G. boninense than B. cepacia UPMB3 (21.27% and 8.89%) compared to control treatment after 7 days of incubation respectively. The findin

Cough is a major symptom of respiratory-related diseases. There exists a tremendous amount of work in detecting coughs from audio but there has been no effort to identify coughs from solely inertial measurement unit (IMU). Coughing causes motion across the whole body and especially on the neck and head. Therefore, head motion data during coughing captured by a head-worn IMU sensor could be leveraged to detect coughs using a template matching algorithm. In time series template matching problems, K-Nearest Neighbors (KNN) combined with elastic distance measurement (esp. Dynamic Time Warping (DTW)) achieves outstanding performance. However, it is often regarded as prohibitively time-consuming. Nearest Centroid Classifier is thereafter proposed. But the accuracy is comprised of only one centroid obtained for each class. Centroid-based Classifier performs clustering and averaging for each cluster, but requires manually setting the number of clusters. We propose a novel self-tuning multi-centroid template-matching algorithm, which can automatically adjust the number of clusters to balance accuracy and inference time. Through experiments conducted on synthetic datasets and a real-world earbud-based cough dataset, we demonstrate the superiority of our proposed algorithm and present the result of cough detection with a single accelerometer sensor on the earbuds platform.

Human affects are complex paradox and an active research domain in affective computing. Affects are traditionally determined through a self-report based psychometric questionnaire or through facial expression recognition. However, few state-of-the-arts pieces of research have shown the possibilities of recognizing human affects from psychophysiological and neurological signals. In this article, electroencephalogram (EEG) signals are used to recognize human affects. The electroencephalogram (EEG) of 100 participants are collected where they are given to watch one-minute video stimuli to induce different affective states. The videos with emotional tags have a variety range of affects including happy, sad, disgust, and peaceful. The experimental stimuli are collected and analyzed intensively. The interrelationship between the EEG signal frequencies and the ratings given by the participants are taken into consideration for classifying affective states. Advanced feature extraction techniques are applied along with the statistical features to prepare a fused feature vector of affective state recognition. Factor analysis methods are also applied to select discriminative features. Finally, several popular supervised machine learning classifier is applied to recognize different affective states from the discriminative feature vector. Based on the experiment, the designed random forest classifier produces 89.06% accuracy in classifying four basic affective states., Comment: 18 pages, 7 figures

Delay Tolerant Network (DTN) architecture comprises of portable devices known as nodes, considered a resource-limited networking system. These nodes in DTN utilize the ‘Store Carry and Forward’ approach to route data since the end to end connections are absent here due to a large number of constant intermittent connectivity. The energy quantity of nodes is restricted because limited-lifetime rechargeable batteries drive them. Accordingly, energy is an essential resource in DTN scenarios. For efficient network performance, including proper energy usage, nodes need to expense a minimum amount of energy. For this reason, it is essential to select an energy-efficient forwarding strategy and exhibit excellent performance among existent forwarding approaches in the DTN environment for routing messages effectively. In this paper, we have studied the energy efficiency of conventional DTN routing protocols: Epidemic, Spray and Wait, Spray and Focus, MaxProp, and PRoPHET on the impact of varying both buffer size and node density. We analyzed their energy consumption and compared their performance based on five performance metrics: average remaining energy, delivery ratio, average delay, transmission cost, and average hop count, respectively. Using ONE simulator, we performed a simulation with varying node density (while buffer size is fixed) and varying buffer size (while node density is fixed). From the outcomes of simulation, we found that Spray and Wait are the most energy-efficient DTN routing protocols. On the contrary, Spray and Focus possessed as the best performer in terms of average hop count, average delay, delivery ratio and transmission cost among conventional DTN routing protocols.

Pseudomonas species have founded as greatest and potentially most promising group of plant growth promoting rhizobacteria (PGPR). Pseudomonas aeruginosa UPMP3 is an important PGPR isolated from oil palm rhizosphere. This rhizobacteria is likely to synthesize and release phytohormone indole-3 acetic acid (IAA). Production of IAA is one of the main reasons to promote plant growth and yield. The aim of this study was to detect, identify and quantify the IAA production by P. aeruginosa UPMP3 in vitro and its influence on oil palm seedling growth. Nutrient broth medium supplemented with 1-5 mg/ml L-tryptophan and without L- tryptophan were used for bacterial culture. The pH levels of culture media were optimized under shaken and static conditions and incubated at 28±2°C in different incubation periods. The production of IAA by P. aeruginosa UPMP3 was extracted, purified, detected and quantified by Thin Layer Chromatography (TLC) and High Performance Liquid Chromatography (HPLC) analyses. Production of IAA was quantified by HPLC in liquid culture and achieved 12.08µg/ml with a retention time of 13.711 min. On the other hand, the maximum 52 µg/ml IAA was recorded in the medium supplemented with 4 mg/ml L- tryptophan in compare to control. The optimum pH level of the culture medium was recorded as 7 under shaken conditions at 150 rpm with 5 days incubation. The influence of IAA produced by the UPMP3 on oil palm seedling growth was carried out in the pot experiment. The germinated oil palm seedlings were treated with the extract of bacterial strain and observed a positive effect on seedling growth in respect to average root and leaf number, root, shoot, and leaf length compare to the synthetic IAA and the control.

Traffic congestion research is on the rise, thanks to urbanization, economic growth, and industrialization. Developed countries invest a lot of research money in collecting traffic data using Radio Frequency Identification (RFID), loop detectors, speed sensors, high-end traffic light, and GPS. However, these processes are expensive, infeasible, and non-scalable for developing countries with numerous non-motorized vehicles, proliferated ride-sharing services, and frequent pedestrians. This paper proposes a novel approach to collect traffic data from Google Map's traffic layer with minimal cost. We have implemented widely used models such as Historical Averages (HA), Support Vector Regression (SVR), Support Vector Regression with Graph (SVR-Graph), Auto-Regressive Integrated Moving Average (ARIMA) to show the efficacy of the collected traffic data in forecasting future congestion. We show that even with these simple models, we could predict the traffic congestion ahead of time. We also demonstrate that the traffic patterns are significantly different between weekdays and weekends.

The bifunctional Ga/HZSM-5 catalyst has been proven having the capability to increase the selectivity of aromatics production during catalytic pyrolysis of furan and woody biomass. However, the reaction chemistry and kinetics of pyrolysis of herbaceous biomass promoted by Ga/HZSM-5 is rarely reported. Pyrolysis–gas chromatography/mass spectrometry (Py–GC/MS) analysis and non-isothermal thermogravimetric analysis at four heating rates were carried out to investigate the decomposition behavior and pyrolysis kinetics of corn stalk without and with Ga/HZSM-5. The effective activation energies for corn stalk pyrolysis were calculated by using the Friedman isoconversional method. The Py–GC/MS analysis results indicated that the Ga/HZSM-5 catalyst had a high selectivity toward producing the aromatic chemicals of xylene, toluene and benzene, whereas the major products from non-catalytic pyrolysis of corn stalk were oxygenated compounds. The presence of Ga/HZSM-5 could significantly reduce the effective activation energies of corn stalk pyrolysis from 159.9–352.4 kJ mol−1 to 41.6–99.8 kJ mol−1 in the conversion range of 0.10–0.85.

As GPS cannot reliably reach and cover any covered area or indoor location, many past studies and research considered the application of Signals of Opportunity (SoOp) such as Wi-Fi, Bluetooth, RFID, Cellular Networks, and FM signals for indoor positioning and object tracking. However, not all SoOp can be part of the indoor positioning experiment due to some limitations; for example, not all signals are cost-effective; some are less accurate in finding position; some require additional hardware for deployment. In this study, a new technique of indoor navigation is introduced exploiting AM radio signals and the received signal strength (RSS) fingerprinting method. This research investigates the fingerprinting technique with the deterministic approach via three algorithms such as Nearest Neighbour (NN), K-Nearest Neighbour (KNN), and K- Weighted Nearest Neighbour (KWNN) for measuring the mean distance error (MDE) as the prime indicator of localization accuracy. The result demonstrates that the Nearest Neighbour algorithm performed best for K=1 and the lowest mean distance error in the ground and the first floor is less than 3m.

The study examines the internal financial structure of ASEAN banking and its efficiency. In doing so, it concentrates on the characteristics of banking sector efficiency and practises within the ASEAN countries banking sectors by using Phillips and Sul (2007a, 2007b and 2007c) approach. Firstly, by applying non-parametric Data Envelopment Analysis (Simar and Wilson, 2007) technique, it considers the technical efficiency of ASEAN banks. The outcome confirms that, at the aggregate level, the ASEAN banking markets are diverging. Further investigation indicates comparatively weak convergence between banking sector of ASEAN though there is still potential of catching up process between them. It can be concluded that ASEAN countries are inter-related and most of the banks are catching up within the different club and different degree of convergence. In terms of efficiency, the results show that at aggregate level, the ASEAN countries are competitive among each other acknowledged by the identification of 3 convergence clubs. The study is able to gain an insight overview regarding ASEAN banking efficiency and its ability to catch up between member countries. These could lead to a successful financial integration, fullfiling the objective to create a potential single currency within ASEAN countries in the near future.

In communication-denied or contested environments, Line-of-Sight (LoS) communication (e.g free space optical communication using infrared or visible light) becomes one of the most reliable and efficient ways to send information between geographically scattered mobile units. In this paper, we consider the problem of planning optimal locations and trajectories for a group of autonomous vehicles to see a set of units that are dispersed in an environment with obstacles. The contributions of the paper are the following: 1) We propose centralized and distributed algorithms to verify that the vehicles and units form a connected network through LoS; 2) We present an algorithm that can maintain visibility-based connectivity, if possible, by relocating a single vehicle; and 3) We study the computational, Comment: Submitted in Robotics and Autonomous Systems

Many robotic systems are remotely operated nowadays that require uninterrupted connection and safe mission planning. Such systems are commonly found in military drones, search and rescue operations, mining robotics, agriculture, and environmental monitoring. Different robotic systems may employ disparate communication modalities such as radio network, visible light communication, satellite, infrared, Wi-Fi. However, in an autonomous mission where the robots are expected to be interconnected, communication constrained environment frequently arises due to the out of range problem or unavailability of the signal. Furthermore, several automated projects (building construction, assembly line) do not guarantee uninterrupted communication, and a safe project plan is required that optimizes collision risks, cost, and duration. In this thesis, we propose four pronged approaches to alleviate some of these issues: 1) Communication aware world mapping; 2) Communication preserving using the Line-of-Sight (LoS); 3) Communication aware safe planning; and 4) Multi-Objective motion planning for navigation. First, we focus on developing a communication aware world map that integrates traditional world models with the planning of multi-robot placement. Our proposed communication map selects the optimal placement of a chain of intermediate relay vehicles in order to maximize communication quality to a remote unit. We also vi propose an algorithm to build a min-Arborescence tree when there are multiple remote units to be served. Second, in communication denied environments, we use Line-of-Sight (LoS) to establish communication between mobile robots, control their movements and relay information to other autonomous units. We formulate and study the complexity of a multi-robot relay network positioning problem and propose approximation algorithms that restore visibility based connectivity through the relocation of one or more robots. Third, we develop a framework to quantify the safety sc

Background: In February 2016, a new fungal disease was spotted in wheat fields across eight districts in Bangladesh. The epidemic spread to an estimated 15,000 hectares, about 16 % of the cultivated wheat area in Bangladesh, with yield losses reaching up to 100 %. Within weeks of the onset of the epidemic, we performed transcriptome sequencing of symptomatic leaf samples collected directly from Bangladeshi fields. Results: Reinoculation of seedlings with strains isolated from infected wheat grains showed wheat blast symptoms on leaves of wheat but not rice. Our phylogenomic and population genomic analyses revealed that the wheat blast outbreak in Bangladesh was most likely caused by a wheat-infecting South American lineage of the blast fungus Magnaporthe oryzae. Conclusion: Our findings suggest that genomic surveillance can be rapidly applied to monitor plant disease outbreaks and provide valuable information regarding the identity and origin of the infectious agent.

Securities and Exchange Commission in Bangladesh set some rules about the Corporate Governance in 2006 and advises the listed company to follow those rules to ensure proper Corporate Governance. The main purpose of this study is to know about the disclosure of checklist of Corporate Governance in the banking sector of Bangladesh. Certain analysis was done to conduct this study like- dichotomous and partial compliance to find out the Overall disclosure Index. Chi-square test and cross tabulation are used to find relation between certain variables. Besides, Correlation and regression was performed to analysis those factors that influences the disclosure of Corporate Governance. Annual report of 2011 for 30 banks enlisted under Dhaka Stock Exchange (DSE) has been used to conduct this study. Certain results like higher proportion of independent directors’ gives higher disclosure of corporate governance and more insider ownership results less corporate governance have been found. Recent scam in banking sector is also the result lack of corporate governance in Bangladesh. This research can be used to minimize that scam in the corporate culture. This study creates values not only to those banks but also to the other stakeholders who are related to this. Investors and creditors decision can be influenced by this study. Keywords:Corporate Governance, Dhaka Stock Exchange (DSE), Securities & Exchange Commission (SEC), Disclosure, Banks etc.

There are still research gaps regarding the effects of key parameters such as water-to-cementitious materials ratio (w/cm), type of binder and pore structure characteristics on the response of concrete to special forms of sulfate attack: physical salt attack (PSA) and thaumasite sulfate attack (TSA). Hence, this study aims at developing an innovative type of concrete: nano-modified fly ash concrete, incorporating various dosages of nano-silica (NS) or nano-alumina (NA) and fly ash, and explores its efficiency in resisting various forms of sulfate attack.

Three cucurbitaceous vegetables viz. sweet gourd (BARI Misti Kumra-1, BARI Misti Kumra-2 and Local Misti Kumra), bitter gourd (BARI Karola-1, Taj Karola-88 and Local Karola) and bottle gourd (BARI Lau-3, BARI Lau -4 and Local Lau) were selected to conduct a research on the incidence and host preference of red pumpkin beetle. In the field experiment at all the three stages, the highest incidence of red pumpkin beetle per plant was observed on sweet gourd and lowest on bitter gourd. The maximum number of beetle was observed on Local Misti Kumra at all the three stages and the lowest number of beetle was found on BARI Karola-1 at seedling stage and on Taj Karola-88 at flowering stage and on Local Karola at fruiting stage in the field. In net cage experiment, the highest and lowest number of beetle was also observed on sweet gourd and bitter gourd respectively. In the net cage among the nine varieties, the maximum number of beetle was observed on Local Misti Kumra. At all the three stages the highest percentage of leaf infestation was found on sweet gourd and the lowest on bitter gourd in the field. Among the nine varieties, the highest percentage of leaf infestation was recorded on Local Misti Kumra at all three stages in the field. The highest food consumption was recorded on sweet gourd among the crops and the highest leaf area damage was found on Local Misti Kumra in net cage experiment among the nine varieties. Among three cucurbit crops the total and daily food consumption was found highest on sweet gourd in the laboratory. The total and daily food consumption was the highest on Local Misti Kumra whereas the lowest was recorded on Local Karola among the nine varieties.

Hevea brasiliensis Muell. Arg. is the major source of commercial natural rubber which accounts for 99% of the world natural rubber production. The narrow genetic base, highly heterozygous nature, the long breeding cycle, low fruit set etc are the major limitations of crop improvement by conventional breeding of rubber tree. The recombinant DNA technology combined with tissue culture technique provides opportunity to introduced novel characters into commercially important crop plants, which cannot be achieved easily by conventional breeding. This study was undertaken to establish embryogenic callus culture and plant regeneration protocol via somatic embryogenesis from potential explants of H. brasiliensis (Clone RRIM 901) and evaluate the potential target tissues to design a genetic transformation protocol through Agrobacterium mediated transformation. In attempt to establish the embryogenic callus and somatic embryos of H. brasiliensis, the potential of zygotic embryo, leaf, cotyledon and root in forming embryogenic callus were examined in the basal MS (Murashige and Skoog) medium supplemented with different auxins and cytokinins at various concentrations. The highest callus formation frequency was observed from zygotic embryo explants, in MS medium containing 2.0 mg/L 2,4- D (90%) followed by NAA (50%), picloram (40%) and Dicamba (20%) after 8 weeks of culture. In response to embryogenic callus induction the highest 50% of the zygotic embryo explants produced embryogenic callus. A higher yield 70% of embryogenic callus was obtained when explants were cultured in MS medium containing 2.0 mg/L 2,4- D and 2.0 mg/L KIN (6-furfuralaminopurin) in presence of 0.5 mg/L NAA. Maximum embryos induction frequency (76.66%) were obtained on modified MS medium supplemented with 1.0 mg/L BAP (6- benzylaminopurine), and 2.0 mg/L GA3 (Gibberellic acid) in presence of 0.1 mg/L NAA after 4 weeks of culture. Frequency of embryo maturation was improved (70%) by inducing amino acid gluta

This study is concerned with the profitability and technical efficiency of the selected Auto Rice Mills (ARMs) in Bangladesh. It has made use of data collected from 35 ARMs from the districts of Naogaon, Dinajpur, Mymensingh, Sherpur and Kustia. A structured questionnaire was used to collect the information. The specific objectives of the study are: i. to estimate cost, return and profitability of the ARMs, ii. to determine overall technical efficiency, pure technical efficiency and scale efficiency of the ARMs and iii. to identify factors affecting operation of ARMs. Farm management method has been applied to determine gross margin, net margin, break-even price and break-even production, while Data Envelopment Analysis (DEA) method is used to determine the technical and scale efficiency of the ARMs. An average rice mill was required to spend US$ 5.31 million per year for its paddy processing operations. The variable cost constituted 98.65% while the fixed cost formed only 1.35% of the total cost. Two most important components of the total costs are for paddy and energy, which constitutes about 92.00% and 4.00% respectively. The total revenue of the rice mills comprises 86.14% of polished rice, 4.31% of bran, 4.78% of broken rice, 1.60% of black/dead rice and 2.90% worth of husk. Gross margin and net margin per year ARM are US$ 1.00 million and US$ 0.93 million respectively. As net margins are positive, the ARMs appear to be sustainable not only in the short-run but also in the long-run. The gross and net margins per metric tonne of polished rice are US$ 79.65 and US$ 73.99 respectively. The break-even price per metric tonne of polished rice was found to be US$ 418 while the average break-even production of polished rice was 12,436 metric tonne suggesting also that the mills were running at profit. The overall technical efficiency score of the rice mills is found to be 0.898 comprising 95.80% of pure technical efficiency and 93.70% of scale efficiency. Pure Technica

The key research problem is to investigate the current newspaper industry and the eventual scope for a free newspaper company for building business and the advertize market in Bangladesh. The main purpose of our research is to present the actual situation of the newspaper industry and all branches of the industry in Bangladesh. Also the goal is to figure out the real situation of the advertisement market which is the most important building block for a free newspaper. We have tried to explore the exact state of the newspaper industry in Bangladesh and if there is any scope or feasibility to start a free newspaper company. We tried to find out the growth of the advertising market in print media especially and also focused on readership tendency among the different age groups. After all the analysis we found Bangladesh has a prospective market for starting a free news paper company in the future.

The impact assessment of a weed biocontrol agent offers insight into the potential of the agent and weed responses in long-term weed management. The impact of the jewel beetle Hedwigiella jureceki Obenberger (Coleoptera: Buprestidae) and its first-generation larvae on the long-pod (LP) and short-pod (SP) forms of cat's claw creeper, Dolichandra unguis-cati (L.) L.G. Lohmann (Bignoniaceae) was studied by initiating one, two, or five pairs (male and female) of adults in a glasshouse experiment. With the higher density of adults, the number of leaves increased significantly, as did the numbers of dead leaves and larval mines, and the percentages of affected leaves (up to 40%), and damaged leaf areas (up to 60%). Plants with the five pairs of adults had fewer leaves and reduced dry weights (e.g., tuber biomass was reduced by 50.4–62.5%). Regardless of adult density, the dry weights of the leaves and tubers were lower in the SP form. Conversely, the LP form had fewer leaves (ca. 17% fewer) and slower shoot growth (mean ± SE = 46.3 ± 2.6 cm per week) than the SP form (66.5 ± 3.8 cm per week). However, the number of tubers was not affected by the plant form and adult density. Larval mines (262 ± 30), pupae (186 ± 17), and adults (144 ± 14) in LP form were higher than in SP form (243 ± 20, 167 ± 14, and 115 ± 11, respectively). Likewise, the average weight of adults that emerged on the LP form was higher. Therefore, H. jureceki impacted both LP and SP forms. The SP form showed a higher impact in terms of reduced tuber growth, and the LP form is anticipated to be harmed more in the long term as a higher number of H. jureceki offspring developed on it when competition for food was increased.

The impact assessment of a weed biocontrol agent offers insight into the potential of the agent and weed responses in long-term weed management. The impact of the jewel beetle Hedwigiella jureceki Obenberger (Coleoptera: Buprestidae) and its first-generation larvae on the long-pod (LP) and short-pod (SP) forms of cat's claw creeper, Dolichandra unguis-cati (L.) L.G. Lohmann (Bignoniaceae) was studied by initiating one, two, or five pairs (male and female) of adults in a glasshouse experiment. With the higher density of adults, the number of leaves increased significantly, as did the numbers of dead leaves and larval mines, and the percentages of affected leaves (up to 40%), and damaged leaf areas (up to 60%). Plants with the five pairs of adults had fewer leaves and reduced dry weights (e.g., tuber biomass was reduced by 50.4–62.5%). Regardless of adult density, the dry weights of the leaves and tubers were lower in the SP form. Conversely, the LP form had fewer leaves (ca. 17% fewer) and slower shoot growth (mean ± SE = 46.3 ± 2.6 cm per week) than the SP form (66.5 ± 3.8 cm per week). However, the number of tubers was not affected by the plant form and adult density. Larval mines (262 ± 30), pupae (186 ± 17), and adults (144 ± 14) in LP form were higher than in SP form (243 ± 20, 167 ± 14, and 115 ± 11, respectively). Likewise, the average weight of adults that emerged on the LP form was higher. Therefore, H. jureceki impacted both LP and SP forms. The SP form showed a higher impact in terms of reduced tuber growth, and the LP form is anticipated to be harmed more in the long term as a higher number of H. jureceki offspring developed on it when competition for food was increased.

This dissertation is a study of the dynamics and direction of contemporary Islamic activism. It examines why some Muslims turn to Islamic activism and what determines the direction of this movement. It focuses on the Jamaat-e-Islami of Bangladesh, one of the most influential Islamic activist movements in South Asia. The study particularly explores the factors that contributed to the rise of the Jamaat-e-Islami in Bangladesh, and the subsequent transformation of this movement. The basic premise of this study is that the appeal of the contemporary Islamic activism is primarily religious, but wherever and whenever it participates in a democratic system, moderation is critical to its wider appeal and political success. By examining the historical roots, ideological discourse, organizational mechanism and the strategy of the Jamaat-e-Islami based on both primary and secondary source materials, the study uncovers that while at the core of this movement is a religious reawakening and rhetoric that were generated by new kind of Islamic discourses and sustained by a well-knit organizational network, this awakening being the result of one particular reading of Islam has attracted only a limited number of adherents. Having failed to win the hearts and minds of the majority as reflected in repeated electoral showings, the Jamaat has turned to redefine its ideology and socio-political agenda by adopting a pragmatic and relatively liberal approach in the political arena. While it is still experiencing dilemmas in reconciling and re-interpreting much of its agenda, the transformation the party has gone through in Bangladesh is significant, for it demonstrates its flexible character and a trend toward further moderation. Empirical findings of this study have wider theoretical implications. First, contemporary Islamic movements are not necessarily fundamentalist, reactive or radical, as they are often portrayed in the literature of this subject. In contrast, this study finds that wh