Small Category Research Articles

Cartesian double theories: A double-categorical framework for categorical doctrines

The categorified theories known as “doctrines” specify a category equipped with extra structure, analogous to how ordinary theories specify a set with extra structure. We introduce a new framework for doctrines based on double category theory. A cartesian double theory is defined to be a small double category with finite products and a model of a cartesian double theory to be a finite product-preserving lax functor out of it. Many familiar categorical structures are models of cartesian double theories, including categories, presheaves, monoidal categories, braided and symmetric monoidal categories, 2-groups, multicategories, and cartesian and cocartesian categories. We show that every cartesian double theory has a unital virtual double category of models, with lax maps between models given by cartesian lax natural transformations, bimodules between models given by cartesian modules, and multicells given by multimodulations. In many cases, the virtual double category of models is representable, hence is a genuine double category. Moreover, when restricted to pseudo maps, every cartesian double theory has a virtual equipment of models, hence an equipment of models in the representable case. Compared with 2-monads, double theories have the advantage of being straightforwardly presentable by generators and relations, as we illustrate through a large number of examples.

Alignment and fusion for adaptive domain nighttime semantic segmentation

In the field of autonomous driving technology, both daytime and nighttime scenes are common. However, due to the poor illumination and difficulty in manual annotation of nighttime images, semantic segmentation of nighttime scenes is more challenging compared to daytime scenes. Therefore, achieving significant progress in nighttime semantic segmentation would greatly enhance the effectiveness of the application of autonomous driving scenarios. In this work, we investigate the problem of domain-adaptive nighttime semantic segmentation (DANSS). The problem aims to learn semantic segmentation in nighttime scenes by leveraging a labeled Cityscapes dataset and unlabeled but roughly aligned day-night image pairs. To address this, we propose an Align and Fusion Network (AAFnet), a network for adaptive domain nighttime semantic segmentation. AAFnet utilizes a novel DAFormer as the backbone to separately compute features for dynamic and static objects during the training process. It also employs methods such as small category sampling and image blending to improve learning effectiveness. Additionally, we propose utilizing local image patches to enhance the results during the testing process. Experimental results demonstrate a significant improvement of 10.4% compared to the previous method, DANNet. Extensive experiments conducted on the Dark Zurich dataset and Nightdriving dataset validate the effectiveness of the proposed approach. Our method outperforms previous backbones and achieves top-ranking results.

Characterizations of some (weak) Grothendieck topologies

Let [Formula: see text] be a small well-powered category which has pullbacks. The aim of this paper is to study and characterize two (weak) topologies on presheaf category [Formula: see text], one of these is constructed by means of a subfunctor of the Yoneda functor which corresponds to an ideal of [Formula: see text], called (weak) ideal topology, and another constructed by a dominion. Naturally, these (weak) topologies of [Formula: see text] introduce two (weak) Grothendieck topologies on [Formula: see text]. To find and study their relations, using the induced presheaf [Formula: see text] given by the fixed dominion [Formula: see text] on [Formula: see text] we construct an action of [Formula: see text] on the subobject classifier [Formula: see text] of [Formula: see text]. Then, we investigate which one of (weak) Grothendieck topologies corresponds one-to-one to the ideals of [Formula: see text] and which one is a sub [Formula: see text]-set of [Formula: see text]. Moreover, among other things, we give some conditions under which the Grothendieck topologies on [Formula: see text] associated to the double negation topology [Formula: see text] and two topologies [Formula: see text] and [Formula: see text] (mentioned above) are sub [Formula: see text]-sets of [Formula: see text].

Filtered bicolimit presentations of locally presentable linear categories, Grothendieck categories and their tensor products

We investigate two different ways of recovering a Grothendieck category as a filtered bicolimit of small categories and the compatibility of both with the tensor product of Grothendieck categories. Firstly, we show that any locally presentable linear category (and in particular any Grothendieck category) can be recovered as the filtered bicolimit of its subcategories of α-presentable objects, with α varying in the family of small regular cardinals. We then prove that the tensor product of locally presentable linear categories (and in particular the tensor product of Grothendieck categories) can be recovered as a fltered bicolimit of the Kelly tensor product of α-cocomplete linear categories of the corresponding subcategories of α-presentable objects. Secondly, we show that one can recover any Grothendieck category as a filtered bicolimit of its linear site presentations. We then prove that the tensor product of Grothendieck categories, in contrast with the first case, cannot be recovered in general as a filtered bicolimit of the tensor product of the corresponding linear sites. Finally, we show how the first presentation can be helpful when computing tensor products of cocontinuous linear functors between Grothendieck categories.

Studi Keragaman dan Habitat Gastropoda di Daerah Litoral Muara Sungai Ketahun Bengkulu Utara

This research aims to identify the abundance, diversity, dominance and evenness of gastropods at the mouth of the Ketahun River, North Bengkulu. The method used is the "purposive" method. The sampling location was divided into 4 (four) stations, namely in the tidal area and on the right and left in the downstream part of the estuary. The sampling technique uses transects, at each station measuring 20m. Gastropod data analysis uses the Shannon Wiener index, Simpson index, and Pielou index. The research results showed that at the mouth of the Ketahun river, 18 species of gastropods were obtained consisting of 7 genera and 3 families. The diversity index shows, a) station I has an index value of 1.81 (medium category); b) station II has an index value of 0.49 (low category); c) station III has an index value of 1.89 (medium category); d) station IV has an index value of 0.82 (low category). The dominance index shows, a) station I has an index value of 0.23 (low category); b) station II has an index value of 0.70 (medium category); c) station III has an index value of 0.19 (low category); d) station IV has an index value of 0.63 (medium category). The evenness index shows, a) station I has an index value of 0.62 (high category); b) station II has an index value of 0.17 (small category); c) station III has an index value of 0.65 (high category); d) station IV has an index value of 0.28 (small category). The most common gastropod found is the genus Clithon with a total of 1369 individuals. Abundance, diversity, dominance and evenness are determined by abiotic parameters and environmental conditions such as substrate and water quality. Keywords: Bengkulu, Gastropods, Ketahun River

An improved focal loss for imbalanced network traffic classification

With the continued progress of the Internet, network security has become an increasingly significant issue that requires constant attention and research. Network traffic classification is a key technology used to detect and prevent malicious network activity, and it has accordingly received increasing attention and research. However, datasets related to malicious network traffic classification often have imbalanced characteristics. In conventional traffic classification problems with multiple categories, the sample size characteristics of small categories are often overlooked. To address this issue, the focal loss function was proposed, which focuses on small samples by modulating the trade-off between the positive and negative samples through two hyperparameters and . This article uses convolutional neural networks (CNN) to tackle traffic classification problem and explore the optimal values of parameters in this application scenario. Additionally, this work proposed a novel weight allocation formula to replace , which allowed small class traffic to obtain higher accuracy.

High-Dose-Rate Yttrium-90 (90Y) Episcleral Plaque Brachytherapy for Iris and Iridociliary Melanoma

PurposeTo describe a pilot study on the use of single-session, high-dose-rate, FDA-cleared, yttrium-90 (90Y) plaque brachytherapy for iris and iridociliary melanoma. DesignA single-center, clinical case series. ParticipantsSix consecutive patients were included in this study. Each was diagnosed with an iris or iridociliary melanoma based on clinical examination with or without biopsy. MethodsEach tumor was staged according to AJCC criteria and received 90Y eye plaque brachytherapy. The main variables were tumor size, patient age, sex, and method of diagnosis (clinical or biopsy). Surgical techniques, treatment durations, and ocular side effects were recorded. Local control was defined as a lack of tumor growth or regression determined by clinical examinations, including slit-lamp and gonio photography, as well as high-frequency ultrasound measurements. Toxicity parameters included acute and short-term corneal/scleral change, anterior segment inflammation, and cataract progression. ResultsHigh-dose-rate 90Y plaque brachytherapy was used to treat small (AJCC cT1) category melanomas. Single-surgery high-dose-rate irradiations were performed under anesthesia. Due to short treatment durations, high-dose-rate 90Y did not require the additional procedures used for low-dose-rate plaque (e.g., sutures, amniotic membrane epicorneal buffering, Gunderson flaps, and second surgeries for plaque removal). Only conjunctival recession was used to avoid normal tissue irradiation. high-dose-rate 90Y treatment durations averaged 8.8 minutes (median 7.9, range 5.8-12.9). High-dose-rate 90Y brachytherapy was associated with no periorbital, corneal (Descemet’s folds), and/or conjunctival edema. There was no acute or short-term anterior uveitis, secondary cataract, scleropathy, radiation retinopathy, maculopathy, or optic neuropathy. Follow-up was a mean of 16.0 (range 12 to 24) months. Evidence of local control included a lack of expansion of tumor borders (n=6, 100%), darkening and/or atrophy of the tumor surface (n=5/6, 83%), and a mean 24.5% reduction in ultrasonographically measured tumor thickness. There were no cases of metastatic disease. ConclusionHigh-dose-rate 90Y brachytherapy allowed for single-surgery, minimally invasive, outpatient irradiation of iris and iridociliary melanomas.

Evaluating the extend of marketing plan in Nigerian business organization

This study assessed the extent to which Nigerian firms planed their marketing programmes and also the extent of formalization of marking planning. It was found that some forms of planning were in operation among many firms. However, only one-third of the firms of studied adopted a comprehensive, formalized, approach, and this was found to be a function of size of operation – large firms tended to do so. We further found that nearly half of the firms prepared forecasts and budgets in lieu of marketing plans, and this was attributed to the intellectual rigour involved in marketing planning. It has been recommended, among other measures that, experienced professionals be always appointed to head the marketing function in business organizations. The population for this study consisted of firms in Lagos employing 50 or more persons engaged in manufacturing or distributive trade. Only such firms listed in the 1986 edition of the Lagos state industrial Directory were included. For the purpose of this study, we classified the firms into categories, A, B, C, as small categories, D and E as medium, and categories F, G, H and I, as large. This implies that we focused attention on categories D to I while ignoring categories A to C. Our concentration on medium-large firms is borne out of our conviction that their size of operations may necessitate the need for elaborate planning.

Some like it burnt: species differences in small mammal assemblage in a Mediterranean basin nearly 3 years after a major fire

The habitat accommodation model (HAM) is a theoretical framework that predicts wildlife community recovery based on their habitat requirements. While post-fire habitat-related research is well documented in the Mediterranean basin, studies specifically focusing on HAM are scarce. Here, we described the small mammal assemblage in a Mediterranean area ~3 years after a fire, specifically examining three functional small mammalian categories: ground-foraging insectivorous, ground-foraging herbivorous/granivorous, and arboreal-foraging species. The study was conducted in Monte Pisano (Italy), where fire burnt ~12 km2 in September 2018. A stratified random sampling was adopted, basing on burnt status and forest type. In each of the 50 sites, during late spring-summer 2021, 12 hair-tubes were deployed, and collected hairs were taxa-attributed based on morphology. A presence/absence dataset was built, and db-RDA was used to explore assemblage composition, and single-species occupancy models to test specific hypotheses. The relative abundance of ground-foraging herbivorous/granivorous was higher in the burnt area, characterised by a dense undergrowth, which could be related to anti-predatory strategies and food opportunities. Insectivorous could be in a recolonisation phase, masking their earlier absence, which could explain why their abundance was not associated with any factor tested. Arboreal-foraging species were associated with forest type, indicating a primary role for tree cover and other factors such as rocky cover and likely in situ survival. The HAM was overall confirmed also in Mediterranean basin ecosystems. This may facilitate predictions about post-fire animal successions, which in turn may provide valuable insights into post-fire management practices and biodiversity conservation strategies.

Crop–livestock-integrated farming system: a strategy to achieve synergy between agricultural production, nutritional security, and environmental sustainability

IntroductionClimate change, nutritional security, land shrinkage, and an increasing human population are the most concerning factors in agriculture, which are further complicated by deteriorating soil health. Among several ways to address these issues, the most prominent and cost-effective means is to adopt an integrated farming system (IFS). Integrating farming systems with livestock enables a way to increase economic yield per unit area per unit of time for farmers in small and marginal categories. This system effectively utilizes the waste materials by recycling them via linking appropriate components, thereby minimizing the pollution caused to the environment. Further integrating livestock components with crops and the production of eggs, meat, and milk leads to nutritional security and stable farmer's income generation. So, there is a dire need to develop an eco-friendly, ecologically safe, and economically profitable IFS model.MethodsAn experiment was conducted to develop a crop–livestock-based integrated farming system model for the benefit of irrigated upland farmers in the semi-arid tropics for increasing productivity, farm income, employment generation, and food and nutritional security through efficient utilization of resources in the farming system.Results and discussionThe IFS model has components, viz., crop (0.85 ha) + horticulture (0.10 ha) + 2 cattles along with 2 calves in dairy (50 m2) + 12 female goats and 1 male goat (50 m2) + 150 numbers of poultry birds (50 m2) + vermicompost (50 m2) + kitchen garden (0.02 ha) + boundary planting + supporting activities (0.01 ha) in a one-hectare area. The model recorded a higher total MEY (162.31 t), gross return (689,773), net return (317,765), and employment generation (475 mandays). Further negative emissions of −15,118 CO2-e (kg) greenhouse gases were recorded under this model. The study conclusively reveals that integration of crop, horticulture, dairy, goat, poultry, vermicompost production, kitchen garden, and boundary planting models increases the net returns, B:C ratio, employment generation, nutritional security, and livelihoods of small and marginal farmers.

KDTM: Multi-Stage Knowledge Distillation Transfer Model for Long-Tailed DGA Detection

As the most commonly used attack strategy by Botnets, the Domain Generation Algorithm (DGA) has strong invisibility and variability. Using deep learning models to detect different families of DGA domain names can improve the network defense ability against hackers. However, this task faces an extremely imbalanced sample size among different DGA categories, which leads to low classification accuracy for small sample categories and even classification failure for some categories. To address this issue, we introduce the long-tailed concept and augment the data of small sample categories by transferring pre-trained knowledge. Firstly, we propose the Data Balanced Review Method (DBRM) to reduce the sample size difference between the categories, thus a relatively balanced dataset for transfer learning is generated. Secondly, we propose the Knowledge Transfer Model (KTM) to enhance the knowledge of the small sample categories. KTM uses a multi-stage transfer to transfer weights from the big sample categories to the small sample categories. Furthermore, we propose the Knowledge Distillation Transfer Model (KDTM) to relieve the catastrophic forgetting problem caused by transfer learning, which adds knowledge distillation loss based on the KTM. The experimental results show that KDTM can significantly improve the classification performance of all categories, especially the small sample categories. It can achieve a state-of-the-art macro average F1 score of 84.5%. The robustness of the KDTM model is verified using three DGA datasets that follow the Pareto distributions.

Agrarian Distress and Indebtedness Among Rural Agricultural Households in Uttar Pradesh

Agriculture is the largest sector of the Indian economy in terms of population dependent on it, where more than 85% of farmers belong to small and marginal categories. In India, agriculture has made commendable progress after the Green Revolution, and with a spectacular rise, the country has become self-sufficient in the case of most of the farm products. Despite it, farmers continued to languish in poverty. As time passes, farming has become less remunerative for most of the farming community, which has led to increasing debt and overall distress. Along with this, the consistent crop failure, coupled with the high cost and vagaries of the market, have made life worse for them. According to the latest data released by ‘National Crime Records Bureau (NCRB) on December 4, 2023, there were 11,290 suicide cases reported by farmers from across the country during 2022, with an increase of 3.7% from 2021. In India, more than half of farmers were in debt and as per NCRB data, there were 7034 farmers’ suicide cases due to indebtedness in 2022. Moreover, the state of Uttar Pradesh saw the highest increase in the number of suicides among all states, with a 42.13% rise when compared with 2021. In this context, the present paper examines agrarian distress and suicide among farmers in the country as well as in the state of Uttar Pradesh. The study, primarily based on secondary sources, tries to explore the underlying reasons for rural indebtedness or bankruptcies, whether it is natural, institutional, social, financial, or government policy-led. The paper also suggests some possible solutions, both short-term and medium-term, including coverage of all crops under Minimum Support Price (MSP); improvement of rural credit delivery system through institutional credit agencies; crop insurance coverage; and provision of input subsidies that can mitigate the agrarian distress among farmers.

The completion of d-abelian categories

Let A be a finite-dimensional algebra, and M be a d-cluster tilting subcategory of mod A. From the viewpoint of higher homological algebra, a natural question to ask is when M induces a d-cluster tilting subcategory in Mod A. In this paper, we investigate this question in a more general form. We consider M as an essentially small d-abelian category, known to be equivalent to a d-cluster tilting subcategory of an abelian category A. The completion of M, denoted by Ind(M), is defined as the universal completion of M with respect to filtered colimits. We explore Ind(M) and demonstrate its equivalence to the full subcategory Ld(M) of ModM, comprising left d-exact functors. Notably, Ind(M) as a subcategory of ModMEff(M) falls short of being a d-cluster tilting subcategory since it satisfies all properties of a d-cluster tilting subcategory except d-rigidity. For a d-cluster tilting subcategory M of mod A, M→ consists of all filtered colimits of objects from M, is a generating-cogenerating, functorially finite subcategory of Mod A. The question of whether M→ is a d-rigid subcategory remains unanswered. However, if it is indeed d-rigid, it qualifies as a d-cluster tilting subcategory. In the case d=2, employing cotorsion theory, we establish that M→ is a 2-cluster tilting subcategory if and only if M is of finite type. Thus, the question regarding whether M→ is a d-cluster tilting subcategory of Mod A appears to be equivalent to Iyama's question about the finiteness of M. Furthermore, for general d, we address the problem and present several equivalent conditions for Iyama's question.

Economic Analysis and Feasibility of Self-propelled Rotor Power Weeder

The manual weeding process is laborious, requires significant effort, and is a costly operation. Farmers typically allocate 30 to 40 percent of their total crop production expenses to the weeding process. The research was carried out by development and economic feasibility of self-propelled weeder for small category farmers. Economic analysis and feasibility of developed mechanical weeder was evaluated by considering cost of operation, breakeven point and payback period. The utilization of a self-propelled weeder for mechanical weeding can result in cost savings of up to 45% compared to the manual weeding method. The estimated break-even point (BEP), calculated based on time and area, indicates potential savings of 209.6 hours and 13.63 ha. The payback period, assessed on a time basis for the rotary power weeder, was determined to be 1.02 years. The total development cost for the self-propelled power weeder was recorded as 8050 Rupees. By adopting self-propelled power weeder, farmer saving 1856 ₹/ha directly over manual method.

Kompetensi Pengguna dan Program Partnership dalam Mempengaruhi Pengambilan Keputusan melalui Implementasi Sistem Informasi Akuntansi

Objectives: This study seeks to validate the impact of user competency and partnership programs on decision-making through the implementation of an Accounting Information System (AIS).Design/method/approach: Employing a quantitative approach, the research focuses on Pekanbaru City Micro, Small, and Medium Enterprises (MSMEs), utilizing purposive sampling for sample selection. Primary data are collected through a questionnaire, and data analysis involves descriptive methods and Partial Last Square (PLS) data analysis.Results/findings: The study reveals that the knowledge and skills possessed by AIS users contribute to enhanced AIS application in MSMEs, facilitating the timely, routine, accurate, and comprehensive generation of financial information or reports aligned with user needs. The competence of AIS users also positively impacts decision-making, enabling MSMEs to establish goals, provide alternatives, minimize uncertainty, and evaluate decisions made. However, the partnership program associated with AIS in Pekanbaru City MSMEs is currently suboptimal, exhibiting no significant effect on AIS implementation.Theoretical contribution: This study explores the novel examination of the impact of a partnership program on decision-making through the implementation of an accounting information system..Practical contribution: The findings indicate that the existing partnership program does not influence AIS implementation. Thus, government or related institutions can enhance similar programs by prioritizing improvements in AIS implementation.Limitations: The study's limitations include the non-classification of MSMEs into micro, small, and medium categories during questionnaire distribution and a sample size comprising only 83%.

The successive dimension, without elegance

Experience shows that the poset of levels (or dimensions) of the topos of presheaves on some elegant Reedy categories may be equipped with a monotone increasing ‘successor’ function which, as the case of simplicial sets illustrates, is different from Lawvere’s Aufhebung in general. We prove that a similar result holds for the topos of presheaves on a small category with split-epi/mono factorizations; a typical feature of categories that are Reedy elegant, or skeletal, or graphic (von Neumann-)regular, but more general. In fact, we show that the more general ‘successor’ may be described as a function on the poset of full subcategories of the site that are closed under subobjects.

Cone complexes and group actions

We introduce the notion of cone complexes to conveniently handle the quotients of simplicial complexes by group actions. Cone complexes are a generalization of simplicial complexes described purely in terms of posets (partially ordered sets). Our main focus of this paper is to study group actions on cone complexes, and compare them with the case of small categories. We clarify how conditions of group actions related to quotients are changed by functors between cone complexes and small categories.

Heavily separable functors of the second kind and applications

We introduce heavily separable functors of the second kind and study them in three different situations. The first of these is with restrictions and extensions of scalars for modules over small preadditive categories. The second is with free functors taking values in Eilenberg-Moore categories associated to a monad or a comonad. Finally, we consider entwined modules and give if and only if conditions for heavy separability of the second kind for functors forgetting either the comodule action or the module action.

An Evaluation of Large Cap, Mid Cap, and Small Cap Mutual Funds: Return, Risk, and Investor Considerations

Investing in mutual funds presents a straightforward entry point into the stock market, particularly for novice investors. Regardless of one’s familiarity with stock markets or age, mutual funds offer accessible investment opportunities. However, selecting the most suitable mutual fund scheme can be challenging, given investors’ varying risk appetites and financial goals. This study aims to assist such investors grappling with scheme selection dilemmas by analysing mutual fund schemes of many types and determining which ones offer superior returns at lower risks. The research evaluates the risk and return profiles of various mutual fund schemes taken up for the study. Utilising various ratio like Jensen’s, Sharpe, and Treynor, present discussion compares the returns by considering the systematic and unsystematic risk of selected schemes within each category. The analysis spans 5 years, from April 01, 2017, to March 31, 2022, drawing data from secondary sources including Moneycontrol, journals, books, and official websites of relevant financial bodies. BNP Paribas Large Cap Fund (G) emerges as a top performer in the sphere of large cap funds according to Sharpe and Jensen’s measures, while Edelweiss Large Cap Fund-A(G) excels by Treynor ratio. Among mid cap schemes, Edelweiss Mid Cap Fund (G) consistently outperforms its peers across all evaluated ratios. In the small cap category, Axis Small Cap Fund-Regular (G) stands out for delivering the highest returns on the basis of Jensen’s, Sharpe and Treynor ratios. Ultimately, this study serves to guide investors 2024in navigating the complexities of mutual fund selection, offering insights into which funds align best with their risk-return preferences and investment objectives. Findings reveal distinct return patterns across fund categories, with specific funds outperforming peers based on various return metrics. The study concludes that while BNP Paribas Large Cap mutual Fund (G) excels in large cap schemes, Edelweiss Mid Cap mutual Fund (G) stands out among mid cap funds, and Axis Small Cap mutual Fund-Regular (G) leads in the small cap mutual funds category. Implications suggest that investors can make informed decisions based on risk-return preferences, with large cap funds offering stability, mid cap funds providing moderate returns, and small cap funds presenting opportunities for better growth with higher risk acceptance capacity.

Geospatial analysis for prediction of potential fishing zones in the Banggai Waters using hydroacoustic data and satellite remote sensing

Banggai waters are a strategic area rich in fish resources and a source of livelihood for most of the local population. Mapping fish stocks is important to ensure the sustainability of fishing in the area. One of them uses hydroacoustic data and satellite imagery to provide more comprehensive results on fish stocks and more accurate predictions of potential fishing zones. This study will spatially estimate fish density in Banggai waters using hydroacoustic data and investigate the distribution of chlorophyll-a and sea surface temperature using Himawari-8 imagery. Hydroacoustic data processing will produce target strength (TS) and volume backscatter strength (Sv) values to determine the value and distribution of fish density. Meanwhile, ZPPI results are obtained from the overlay of chlorophyll-a and sea surface temperature. From the results of mapping the potential fishing zone, the high category has the highest percentage of 43%, low 28% and medium 29%. The ZPPI point distribution map then shows the concentration of ZPPI points in the southern Banggai waters. The fish density distribution map shows that the spatial distribution tends to be around the coast of Banggai Peninsula, dominated by small and very small fish categories. Correlation of fish density point distribution with ZPPI results and oceanographic parameters has a positive influence on fish density from GAM modelling.