Emotion Recognition using a Multi-view learning method

The increasing use of computing devices and applications in human daily life triggers the need for natural human-computer interaction. Emotion Recognition using multiple features using a semi-serial fusion method is proposed. The study analyses the impact of the feature combinations in the enhancement of the recognition enhancement. The paper presents the use of the multi-view learning principle to a fusion of different features for one emotion expression-based recognition. The results prove that a planned method is operative. The proposed combination method outperforms the use of one type of features and the concatenated way in recognition accuracy, improvement of execution time, and stability.

Mobile Devices in supporting members’ participation to Village Community Banks: The Design Thinking Approach

Village Community Banks (VICOBA) and mobile phones have become assistance to many rural societies in encountering financial problems. However, mobile phone usages to VICOBA are isolated resulting in unsubstantial benefits that do not foster member’s participation. The weak participation of members to Village Community Banks activities has been one of the major impediments to their success despite owning of mobile devices to some members. The paper applies design thinking approach to explore the situation through identifying the deficiency available in mobile devices to VICOBA microcredit programmes in enhancing member’s participation and suggest the effective usage pattern The study captured 7 groups of village community banks at Ndungu village with a total of 70 members whereas 48 female and 22 male participants. The study employed interview, observation and focus group discussion for data collection. Content analysis and coding approaches were used to analyse data and interpreting findings. The findings observed that there is a big potential in mobile phone use to village community banks activities and explicitly linking between design thinking and the qualitative data collection methods in the inception phase of the design process has been perceived to exist

A Security Framework for Mobile Application Systems: Case of Android Applications.

Smartphones have become a major part of human’s life. And also it has been seen that new mobile applications are built day by day. Currently mobile applications are playing major role in many areas such as banking, social networking, financial apps, entertainment and many more. With this increasing number of applications, security is an important issue. The growth of android market has increased security risk and thus focus should be given to the security. Security is the biggest issue in the field of mobile technology. Therefore, mobile applications need to be assessed and ensure that secure coding practices have been followed during development. Mobile application security breach can lead to fraudulent transactions through mobile applications, confidentiality and revenue loss through communications services misuse. Data that is shared on an unsecured channel is vulnerable to attacks and to stop unauthorized access to this data, there is need to encrypt the data before it is sent to the server. In this research work, different cryptographic algorithms for encrypting data and secure data sharing in mobile applications across communications channels were examined. Simulation methodology was used to investigate a suitable cryptographic algorithm and to design a security framework for mobile applications to solve mobile application security problems. The proposed framework employs the use of Advanced Encryption Standard (AES) algorithm for encrypting meter readings data being exchanged between a smart phone and the server. The results obtained from the simulation of the security framework, showed that the four fields namely: Account number, image path, meter number and phone number on which AES encryption was applied were in an unreadable format (ciphertext), implying that the fields have been successfully encrypted. This solution allows application users to transfer (upload readings) data between a smart phone and database server in a secure manner without facing the problem of data attack. Data being uploaded to the server is encrypted before it is transferred and decrypted once it reaches the server side. This solution addresses android application security in the application and network communications layers and data transmission. The research paper ensures information security is guaranteed between an organisation and its customers.

Feature Selection using Genetic Programming

Extracting useful and novel information from the large amount of collected data has become a necessity for corporations wishing to maintain a competitive advantage. One of the biggest issues in handling these significantly large datasets is the curse of dimensionality. As the dimension of the data increases, the performance of the data mining algorithms employed to mine the data deteriorates. This deterioration is mainly caused by the large search space created as a result of having irrelevant, noisy and redundant features in the data. Feature selection is one of the various techniques that can be used to remove these unnecessary features. Feature selection consequently reduces the dimension of the data as well as the search space which in turn increases the efficiency and the accuracy of the mining algorithms. In this paper, we investigate the ability of Genetic Programming (GP), an evolutionary algorithm searching strategy capable of automatically finding solutions in complex and large search spaces, to perform feature selection. We implement a basic GP algorithm and perform feature selection on 5 benchmark classification datasets from UCI repository. To test the competitiveness and feasibility of the GP approach, we examine the classification performance of four classifiers namely J48, Naives Bayes, PART, and Random Forests using the GP selected features, all the original features and the features selected by the other commonly used feature selection techniques i.e. principal component analysis, information gain, relief-f and cfs. The experimental results show that not only does GP select a smaller set of features from the original features, classifiers using GP selected features achieve a better classification performance than using all the original features. Furthermore, compared to the other well-known feature selection techniques, GP achieves very competitive results.

Water Utility Mobile Application For Meter Reading: Android Based

Water utility companies have a billing process which involves reading the meter as well as knowing the consumption of its customers. The major challenge faced by water utilities in Zambia is inaccurate customer billing. Affecting hugely this inaccurate customer billing is the manual meter reading process. For most water utilities in Zambia meter reading is a very complex task. The current method for meter reading uses manual work of reading the meter and updating details of the customers. The readings are maintained in a record book and due to manual processing of data there are possibilities of generating an incorrect bill. This process cause extra work load on person who is capturing meter readings. The manual process is associated with the following challenges; transfer errors, poor handwriting, manipulation of figures, lack of analysis of readings, low productivity, time pressure, increase in non-revenue water and meter readers not visiting the property. The other problem with manual meter reading is that both the customer and the water utility company have challenges in monitoring and analyzing of consumption trends for customers and it is very difficult to monitor meter readers in the field. Although technologically advanced solutions exist, most water utilities in Zambia have no budget for them. The proposed model comprised of the android mobile application for capturing readings in the field and the web application for analysing readings and monitoring of field workers in real time. A usability test was undertaken consisting of 36 respondents which comprised of 75% meter readers and 25% customers. The results obtained from the mobile application framework, showed that the capturing, uploading, monitoring and analysis of readings were very good for the users and customers who participated in the study. Three user experience areas were used to undertake the survey namely, perceived ease-of-use, perceived flexibility and perceived information accuracy. With the android and web application customers are also able to take their own readings and view graphical usage at any time anywhere.

Exploring Learning Cultures of Digital Immigrants in Technologically mediated Postgraduate distance learning mode at the University of Zambia

In this study, we interrogate the Learning Cultures of Peace Leadership and Conflict Resolution Postgraduate learners in their quest for higher education within the University of Zambia using the Learning Management System. The thrust of the study is on learning cultures of digital immigrants in technologically mediated postgraduate distance learning mode to inform the development of learner responsive services. The study rides on Hermeneutics Phenomenology approach to elicit lived experiences of seventeen (17) purposively selected year two students. The study approach is well suited as it empowers Digital immigrants to voice out their lived realities while following the digital mediated Masters of Science in Peace Leadership and Conflict Resolution Programme to forge best ways of harnessing their learning opportunities. The findings reveal that whereas learners have migrated to the Learning Management System, their learning culture is profoundly still in the print age as they keep on requesting for print based educational resources. Secondly, there is limited interactivity among learners and between learners and their learning facilitators, negating the very essence upon which the Learning Management System has been created. This state of affairs has had a bearing on their motivation to engage effectively in the actual learning process as informed by Moore’s theory of Interactivity and Siemens’ theory of Connectivism. To this end, the study recommends that learners are effectively oriented in the use of digital resources to empower them as they exploit the available educational opportunities. Additionally, there is need to strengthen capacity building mechanisms in order to bridge the gap between the learners and the learning facilitators.

Examining Factors Influencing the Adoption mobile application in collaborative learning

Nations are making huge investments to ensure that their citizens receive the education so that they can contribute to different aspects of development of a country. Despite the investments in education in many countries, researches show that there is need to improve the quality of education offered in tertiary institutions. To improve the quality of education offered institution of higher learning need to consider approaches that place students at the center of the learning process. One such approach is to use mobile application-aided collaborative learning. Mobile application-aided Collaborative learning promises great benefits to students and institutions. However, to successfully implement mobile application aided collaborative learning, it is important to determine if students are ready to use it. In this paper, we use the Unified Theory of Acceptance and Use of Technology (UTAUT) to determine the key factors that should be put in place for collaborative learning to be used successfully. The main contribution that this paper makes is that, it identifies the factors that Copperbelt University needs to work on so that collaborative learning can be used more effectively.

Technological Paradox of Hackers Begetting Hackers: A Case of Ethical and Unethical Hackers and their Subtle Tools

Computer crimes have been in existence for a long time now and hacking is just another way or tool that hackers are now using to perpetrate crime in different form. Hackers Beget Ethical Hackers. A number of people have suffered the consequences of hacker actions. We need to know who these hackers are. We need to know why these hackers exist because hackers have been there and will be there and we can be victims of their existence. In essence hackers seem to beget hackers and the tools that they use are getting more and more advanced by the day. We shall take a quick analysis of selected tools from thousands of tools used by ethical and unethical hackers.We shall systematically review three major types of hackers that we can identify. It is not easy to draw a line between them. Three main hackers and minor hackers have been discussed in this paper. The three main hackers are black hat, grey hat and white hat hackers.We have adopted a systematic review of literature to discuss and analyse some of the common tools the black hat hackers have developed to hack into selected systems and commercial software and why they do it?

Machine Learning Algorithms for automated Image Capture and Identification of Fall Armyworm (FAW) Moths

Automated entomology is one of the field that has received a fair attention from the computer scientists and its support disciplines. This can further be confirmed by the recent attention that the Fall Armyworm (FAW) (Spodoptera frugiperda) has received in Africa particularly the Southern African Development Community (SADC). As the FAW is known for its devastating effects, stakeholders such as the Food and Agriculture Organization (FAO), SADC and University of Zambia (UNZA) have agreed to develop robust early monitoring and warning system. To supplement the stakeholders’ efforts, we choose a branch of artificial intelligence that employs deep neural network architectures known as Google TensorFlow. It is an advanced state-of-the-art machine learning technique that can be used to identify the FAW moths. In this paper, we use Google TensorFlow, an open source deep learning software library for defining, training and deploying machine learning models. We use the transfer learning technique to retrain the Inception v3 model in TensorFlow on the insect dataset, which reduces the training time and improve the accuracy of FAW moth identification. Our retrained model achieves a train accuracy of 57 – 60 %, cross entropy of 65 – 70% and validation accuracy of

Design and Control of a Grid-connected Seven Level Inverter for Photovoltaic Applications

This paper presents the design and closed-loop current control of a grid connected seven-level, 3-phase diode-clamped multilevel inverter for Photovoltaic (PV) applications. The proposed closed loop current control technique is based on the voltage-oriented proportional integral (PI) controller theory. The modulation technique used is level-shifted-carrier sinusoidal pulse width modulation (SPWM). The gain values of PI controller were selected to achieve good current quality and dynamic response. Grid synchronization was achieved by using a synchronous-reference frame phase-locked loop (SRF-PLL). Matlab/Simulink was used for the control system design and simulation. The simulation results show that a 1.34% total harmonic distortion (THD) of the output current was achieved which is within the allowable current distortion limits by international standards. The stability of the system was analyzed using pole-zero mapping and root locus.