Archives-June... 

Climate change and climate variability and financial institutions’ unwillingness to give loans have resulted in many farmers losing confidence in dry land agriculture. Traditional crop insurance methods have also presented challenges due to the risk related to adverse selection and moral hazard resulting in high transaction costs for individual assessment. This study focused on developing a weather index based insurance model that uses artificial neural networks to estimate potential evapotranspiration (ETO) and consequently yield reduction due to moisture stress. Weather data from 2012 to 2015 for Kutsaga area in Harare was used for the study. Seasonal weather data were used as input data to the first model to predict the ETo. The output ETO and effective rainfall data together with the crop factor (Kc), yield reduction factor (Ky), root zone depth (RzD) and root zone moisture (RzM)were used as input data for the second network to compute % yield reduction. Data for maize for the 2012-13 growing season was used for training the network and validating the estimated ETO and % yield reduction. The estimated ETO compared very well with the calculated values with R2 values of above 0.84. The estimated yield reduction % indicated even high accuracies with R2 values of above 0.91. The 2014-2015 growing season resulted in crop loss due to mid-season dry spells and the model predicted a 100% crop loss which means the farmer had to be compensated for the value equivalent to cost of inputs. The model has got potential to be used by insurance companies using weather based data and, with mobile banking transaction costs can be reduced.Climate change and climate variability and financial institutions’ unwillingness to give loans have resulted in many farmers losing confidence in dry land agriculture. Traditional crop insurance methods have also presented challenges due to the risk related to adverse selection and moral hazard resulting in high transaction costs for individual assessment. This study focused on developing a weather index based insurance model that uses artificial neural networks to estimate potential evapotranspiration (ETO) and consequently yield reduction due to moisture stress. Weather data from 2012 to 2015 for Kutsaga area in Harare was used for the study. Seasonal weather data were used as input data to the first model to predict the ETo. The output ETO and effective rainfall data together with the crop factor (Kc), yield reduction factor (Ky), root zone depth (RzD) and root zone moisture (RzM)were used as input data for the second network to compute % yield reduction. Data for maize for the 2012-13 growing season was used for training the network and validating the estimated ETO and % yield reduction. The estimated ETO compared very well with the calculated values with R2 values of above 0.84. The estimated yield reduction % indicated even high accuracies with R2 values of above 0.91. The 2014-2015 growing season resulted in crop loss due to mid-season dry spells and the model predicted a 100% crop loss which means the farmer had to be compensated for the value equivalent to cost of inputs. The model has got potential to be used by insurance companies using weather based data and, with mobile banking transaction costs can be reduced.

Cystic parotid lymphangiomas are benign vascular tumors that rarely appear in adults; they are usually detected in the first two years of life. Their appearance is due to an abnormal development of the lymphatic system during embryogenesis. Herein we report on two cases of intra-parotid cystic lymphangioma that occurred in teen age. A six year girl and a young girl of 13 years age were admitted in our unit because of an indolent and renitent isolated mass occupying the right parotid region. The diagnostic of an intra-parotid cystic lymphangioma was clinically advocated on echographic and CT imaging findings and then definitely proven after histopathologic studies of the ablation fragments. A total parotidectomy was made, removing the entire tumor. The parotid localization of cystic lymphangioma makes their surgical management difficult because of the development of lymphangioma between the branches of the facial nerve. The question of the benefit / risk ratio makes the therapeutic decision more complex because the aggressiveness of the radical tumor resection encounters a benign tumoral origin. It opposes two fundamental principles, first that of being radical on the tumor which is very recurrent and on the principle of remaining functional while preserving the function of the facial nerve.

Organizations came to the realization that it is not economically feasible to maintain the increasing maintenance costs of maintaining data centers with large floor space. One of the main concerns associated with Cloud Computing is the security of the service is that sensitive and important data is no longer only under the control of the data owner as the service provider also has access to that data. In terms of the methodological approach used by this thesis, this thesis will utilize the use of the qualitative method in answering the research questions posed by this thesis. The systematic literature review has shown that there are novel and innovative ways for organizations to mitigate the inherent risks associated with the use of cloud computing.

In order to better understand the zootechnical and economic characteristics of piglet farming, a diagnostic study was conducted from October to December 2019, three months with 30 pig farmers, in the form of cross-sectional and retrospective surveys in the three municipalities of the town of Kindu in the Democratic Republic of Congo. The results showed that pig farming was practiced by both married and educated men and women. They were mainly zimba, Kusu and Nande and associated livestock with agriculture, private activity or trade. They operated in two different livestock systems (traditional and semi-intensive), with a clear predominance of the traditional system (85.5% of farms). The latter was characterized by semi-intensive hog barns (51.1%), where local-bred pigs were mostly reared (90.4% of farms). The average herd size was 19.2 - 20.1 pigs. In 97% of cases, farmers fed the animals from kitchen waste and agricultural and agro-industrial by-products. The average breeding age was 7.6 -1.8 months and the average litter size was 7.4 - 2.2 piglets. This livestock subsector has real potential for poverty reduction because it generates net incomes on average of $40 per piglet, depending on the system. Nevertheless, the lack of training of herders, inadequate pig habitats, food and health constraints remain the main factors hindering the development of pig farming in this city. It would be useful to find alternative ways to improve the habitat conditions, feeding and health monitoring of pigs, accompanied by better organisation and capacity building for producers and other players in the hog sector.

Despite the fact that men constitute a large portion of Nigerian media coverage, it appears that we know little about how they are represented in Nigerian newspapers especially by female writers/journalists. Most studies concerned with gender portrayal in Nigerian media tend to focus on how women are represented mainly by male writers/journalists (e.g. Odejide 1996, Daniel 2011, Adewoye et al 2014). This paper examines the many complex, subtle and contradictory patterns of linguistic representation of men in female-authored articles in Nigerian newspapers. It does so by employing methods of and insights from critical discourse analysis and corpus linguistics tools to analyze 200 female-authored articles on gender and other related issues published in five Nigerian newspapers namely; The Guardian, Vanguard, The Punch, Daily Trust and New Nigerian from 1999 to 2014, a period that has witnessed an unprecedented surge in the wave of gendered discourses in Nigerian print news media. Using Lawrence Anthony’s AntConc version 3.4.3w concordance package, collocational profile of the node word MEN (and its singular variant) was extracted and investigated, and results obtained show that Nigerian female authors generally depict men as both the “problem” and the “answer.” This is evidenced by the prevalence of discourses that construct men as oppressors on one hand, and providers and leaders on the other. These contradictory patterns of representation are understood in this paper as examples of “textual heterogeneity” as exemplified by the range of different identities women have constructed for men. Moreover, these patterns of representations can also be seen as both reinforcing and resisting patriarchal social order existing in our society.

Normally, in the heart of a fetus, there is a ductus arteriosus (DA) which is to close after birth at most three months. Because DA closes, it becomes ligamentum arteriosum. When DA does not close after birth, it is known as patent (open) ductus arteriosus (PDA). The left PDA occurs more than the right PDA. The size of the PDA determines the severity of the condition. Should be diagnosed and treated early in life. It can be treated either with drugs or surgery. When left untreated for a long time, it turns into Eisenmenger syndrome.

The need of automatic classification of text has become very crucial with the rampant growth in the amount of information available in today’s world of fast digitization. Text classification (also called text categorization) is the process of classifying documents into pre-defined categories. A number of techniques have been devised over the ages for text classification. One such technique for text categorization is Support Vector Machine (SVM). SVMs are binary classifiers which classifies text into exactly two categories. Many researchers have found that SVM works reasonably well. In this paper, this technique is used for classifying a set of documents. Five models have been created from five different domains and data from each domain are classified using these models and the accuracy is recorded.

A new hybrid approach with Genetic Algorithm and Simulated Annealing is proposed to solve Multi Job-Shop Scheduling problem, which is one of the well-known hardest combinatorial optimization problems. The main objective of multi job shop scheduling problem is to find a schedule of operations of each job in a set of jobs that can minimize the maximum completion time called makespan. To improve the makespan, SA algorithm has been designed and combined with genetic algorithm . Thus, hybrid GA is implemented over MJSSP and the effectiveness and efficiency is proved by getting promising results for different benchmark job-shop scheduling problem instances.

Malignant triton tumor (MTT) is an extremely rare variant of malignant peripheral nerve sheath tumor (MPNST) with rhabdomyo-sarcomatous differentiation. We report a case of a 14-year-old male without history of neurofibromatosis presented for post-traumatic abdominal pain. Ultrasonography and contrast-enhanced computed tomography (CT) revealed a large heterogeneous intraperitoneal mass near the vascular structures and repressing the adjacent organs. Surgery and anatomopathological examination revealed a malignant triton tumor. MTTs are highly aggressive tumors that are fast-growing with a tendency to recur locally and metastasize early. To date, there is no treatment consensus available yet and the overall prognosis remains dismal.

PPP contracts are contracts with special nature that tend to have longer contract durations compared to other conventional procurement methods; hence, contract re-equilibrium becomes an inevitable need. The objective of the re-equilibrium process is to maintain the IRR of the private sector constant. In this paper, a framework is established to calculate the different re-equilibrium scenarios. The framework proposed in this paper, in addition to addressing a constant IRR, can also be customized to work with any other contractual financial ratio such as ROE. A spreadsheet model is developed with the aid of Microsoft Excel and VBA programming language.