Ethiopia has a big network of water resources comprising lakes, rivers, dams, reservoirs, wetlands and other small water bodies. One of those lakes is Lake Rudolf/Turkana, which is a shared desert lake in Ethiopia and Kenya, the major water discharge around 90% is from Ethiopia through Omo river. The omo Turkana valley is known in its very big fish production especially in Nile perch and Tilapia. This study aimed to assess the fish post-harvest loss, handling, processing and Marketing situations of Lake Rudolf/Turkana Ethiopian side at Dassench woreda where intensive fishing activities under taken. The study uses purposive sampling method with focused group discussion and with field observation as well as secondary data collection. Qualitative and quantitative data were analyzed using SPSS version 26 and descriptive narration method was used to finalize this paper. \nBased on the finding currently Lake Rudolf/Turkana carrying high fishing pressure. Ethiopian and Kenyan fishermen scramble the fish resources. The Ethiopian side Lake Turkana annual fish production is around 517.58 metric tons, where 327.73 metric tons Nile perch and 189.85 Metric tons Tilapia. Together with this the fish post-harvest loss on average is 11%; in which 13% of tilapia fish and 9.85 % Nile perch loss. This loss includes both the physical and quality loss during harvesting, handling (on board, unloading, transporting), processing, distribution, storage and marketing. Therefore, there should be good fish hygienic processing system in place and also studying the microbiological condition of the fish and fish products is critical for public health. Awareness creation to fishermen and development agents is vital and also provision of preservation technologies and equipment’s including cold chains and other processing utilities, development of fish processing shade will improve the quality of the fish products and at the same time reduce the post-harvest loss.

Encryption is the process of scrambling a message\nand can provide a means of securing information. Information\nSecurity is becoming more important in data storage and\ntransmission. It is known that most encryption method in the\nliterature. In this paper we propose a new encryption scheme by\nusing cyclic codes. Our method is based on One Time Pad system.\nWe use the properties of cyclic codes to provide its security.

Introduction: The frequency of delirium among patients with cancer presenting to the emergency department is still unknown and not often recognized. In this study we intend to assess delirium for the cancer patients admitted to ED with the complaint of altered level of consciousness, based on Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) diagnostic criteria, research for influencing etiologic factors and comparison of the brief Confusion Rating Method (bCAM), Mini Mental State Examination (MMSE) and New Delirium Rating Scale (NDRS), which are considered as delirium screening tests. \nMethods: This was a prospective observational study. Richmond Agitation-Sedation Scales (RASS) calculated for all of the patients before applying bCAM. Delirium evaluation could not be made when the RASS score was -4 or -5. The patients with the RASS score between -3 and +4 had been evaluated with bCAM. Delirium was diagnosed when the third or fourth characteristic was positive as well as the first two. The MMSE and NDRS scores of all patients and the duration of three tests were calculated.\nResults: Total 195 patients were included; 112 (57,4%) were male. MMSE and NDRS scores of 26 patients with delirium were 13,46±3,78 (7-20 and 21,42±3,28 (11-26) respectiveley. Harmony between bCAM and MMSE are also statistically significant (Eta=0,70). Application period of bCAM was the shortest as 46,92±6,16 (30-60) sec.\nConclusion: bCAM was applied in the shortest period of time. This result is very useful for the EDs which are racing against time in our country and in the world.

Right now, the 2019 Coronavirus (COVID-19) disease is a significant challenge for health practitioners all over the world. Distinguishing COVID-19 from X-Radiation (X-Ray) images is urgent for determination, evaluation, and treatment. In the present work, we are utilizing artificial intelligence (AI) methods for quick and accurate diagnosis of COVID-19 cases from X-Ray & computerized tomography (CT) images utilizing convolution neural networks (CNNs) resulting in deep transfer learning (DTL). We are trying to advance the symptomatic achievement of combined efforts of human-machine interaction. This is vital in separating COVID-19 patients from those without the infection, where the expense of a mistake is devastating. In this work, we have developed the Inception V3 model (with the depth of 48 layers), VGG-16 model (Visual Geometry Group from Oxford & also called as OxfordNet with the depth of 16 layers), VGG-19 (with the depth of 19 layers) methods for COVID-19 detection. The solvers used Adam, Stochastic Gradient Descent (SGD) method and Limited memory-Broyden–Fletcher–Goldfarb–Shanno Bound constraint method (L-BFGS-B) varying with various activation such as tanh, identity, Logistic and Rectified linear unit (ReLu). The proposed framework has been implemented to train and validate on two datasets. The disease prediction accuracy for (COVID-19 chest xray, 2020) i.e. D1 is: Inception V3 (by Adam as 79.85%, SGD as 80.225%, L-BFGS-B as 83.2%), VGG-16 (by Adam as 77.875%, SGD as 77.2 %, L-BFGS-B as 82.2%), VGG-19 (by Adam as 78.875 %, SGD as 79.875 %, L-BFGS-B as 83.2%). Similarly, the prediction accuracy on the (Talo, M., 2020) i.e. D2 is: Inception V3 (by Adam as 77.9%, SGD as 74.75%, L-BFGS-B as 78.82%), VGG-16 (by Adam as 80.62%, SGD as 78.45%, L-BFGS-B as 80.77%), VGG-19 (by Adam as 79.97%, SGD as 78.02%, L-BFGS-B as 80.25%).

Today, Airbnb is changing the business models of the hospitality industry globally. While a bigger impact has been given by the Airbnb community to the development of the local economy of each country, very little effort has been exerted to investigate the Airbnb pricing issue. This paper provides a report of a study that focuses on modeling hospitality price prediction of the Airbnb dataset with machine learning techniques. Additionally, it is written with an attempt to present the fundamental design and implementation of the machine learning model based on auto-parameters tuning in predicting the Airbnb price. From the results, it has been observed that the two regressor algorithms, namely the Decision Tree and the Random Forest, have been very promising to produce a high percentage of accuracy score with auto-parameters tuning approach.

Cette étude analyse le discours public des populations locales exprimé dans le conflit lié à l’exploitation du site minier de Méa en Nouvelle-Calédonie. Ces populations issues de tribus kanaks, qui pour pouvoir réclamer leur droit à la consultation préalable sur l’usage du territoire, étayent leur discours sur un double mécanisme d’identité sociale et de représentation sociales. Depuis quelques années, différentes initiatives ont été prises, parfois en collaboration avec des ONG, pour faire entendre leur voix : participation à des reportages audiovisuelles, réalisation d’articles de presses, alimentation de pages Facebook... autant de matériaux constituant des corpus lexicaux pertinents pour étudier la manière dont se construit un conflit de représentations sociales liées à l’environnement et à la modification des espaces forestiers. \nEtant entendu que les acteurs de la production de discours sur un territoire local compris en tant qu’objet-enjeu se basent sur un ensemble de représentations qui participent à la régulation des conflits intergroupes, et qui s’insèrent dans le processus de prise de décisions dans le cadre d’une gestion durable de l’environnement. Et qu’il est désormais admis que le territoire est à la fois une condition écologique, un construit social sujet aux jugements de valeur et un phénomène politique polarisant, l’analyse discursive de ce conflit nous permet de concilier la théorie des représentations sociales avec celle du cadrage afin de pouvoir dépasser le concept de résonnance culturelle. Nous montrons ainsi que le concept d’utilisation non voulue de territoire locale (LULU) gagnerait en profondeur en articulant celui-ci avec la théorie des représentations sociales et la théorie des cadres. Cette étude démontre l’importance de la prise en compte de l’expression discursive des représentations sociales dans l’émergence du consensus au sein d’un conflit territorial et communicationnel.