Early disease identification in crops is critical for food security, especially in Sub-Saharan Africa. To identify cassava diseases, professionals visually score the plants by looking for disease indicators on the leaves which is notoriously subjective. Automating the detection and classification of crop diseases could help professionals diagnose diseases more accurately and allow farmers in remote locations to monitor their crops without the help of specialists. Machine learning algorithms have been used in the early detection and classification of crop diseases. However, traditional machine learning algorithms are not calibrated even though they have high accuracy. The ability to provide well-calibrated posterior distributions is one of the most appealing properties of Gaussian processes. Motivated by the current developments in the field of Gaussian Processes, this study proposed a deep Gaussian convolutional neural network model (DGCNN) for the detection and classification of cassava diseases using spectral data. The proposed model uses a hybrid kernel function that is the product of a rational quadratic kernel and a squared exponential kernel. Experimental results revealed that our proposed hybrid kernel function performed better in terms of accuracy of 90.10% when compared to both the squared exponential kernel with an accuracy of 88.01% and the rational quadratic kernel with an accuracy of 88.52%. In our future work, we propose to integrate the Optimised model proposed in this study with the transfer learning approach, a move that may help to improve the model performance.

Computer worm detection has been a challenging and often elusive task. This is partly because of the difficulty of accurately modeling either the normal behavior of computer networks or the malicious actions of computer worms. This paper presents a literature review on the worm detection techniques, highlighting the worm characteristics leveraged for detection and the limitations of the various detection techniques. The paper broadly categorizes the worm detection approaches into content signature based detection, polymorphic worm detection, anomaly based detection, and behavioral signature based detection. The gap in the literature in the techniques is indicated and is the main contribution of the paper.

The process of moving from experimental data to modeling and characterizing the dynamics and interactions in natural processes is a challenging task. This paper proposes an interactive platform for fitting data derived from experiments to mathematical expressions and carrying out spatial visualization. The platform is designed using a component-based software architectural approach, implemented in R and the Java programming languages. It uses experimental data as input for model fitting, then applies the obtained model at the landscape level via a spatial temperature grid data to yield regional and continental maps. Different modules and functionalities of the tool are presented with a case study, in which the tool is used to establish a temperature-dependent virulence model and map the potential zone of efficacy of a fungal-based biopesticide. The decision support system (DSS) was developed in generic form, and it can be used by anyone interested in fitting mathematical equations to experimental data collected following the described protocol and, depending on the type of investigation, it offers the possibility of projecting the model at the landscape level.

This study explored moderating role of entrepreneurial orientation on the relationship between Information Technology Competence and firm performance in Kenya. The impact of IT on FP remains debatable to-date because some results of previous studies have had high variations resulting from diversities in the conceptualization of the key constructs and their interrelationship, coupled with the exclusion of intangible effect of IT on performance. In Kenya, SMEs employ about 85 percent of the workforce. The need to link ITC with FP has become vital for firms striving to achieve superior performance. However, limited attention has been paid to the link and more so to the moderating role of EO on ITC- FP relationship model. To better understand this relationship, this paper adopted a mixed methods research guided by cross-sectional survey design. Quantitative and qualitative techniques were employed to analyze the collected data using SPSS, Ms-Excel, AMOS, SmartPLS, STATA, R-GUI and ATLAS.ti analytical softwares. Analyses were conducted using a two-phase process consisting of CFA and SEM models. The theoretical models and hypotheses were tested based on empirical data gathered from 94 SMEs in the 2013 Top 100 Survey. The study found that ITC had a positive relationship with FP. The results also revealed that EO did not significantly moderate the relationship between ITC and FP in Kenya. However, when run with the interaction term, the Technical (ITC and ISRA)*EO was statistically significant at 10% α-level. This study will enhance the skill set in Kenyan SMEs and produce a more sustainable solution.

Sentiment analysis has become an important area of research in natural language processing. This technique has a wide range of applications, such as comprehending user preferences in ecommerce feedback portals, politics, and in governance. However, accurate sentiment analysis requires robust text representation techniques that can convert words into precise vectors that represent the input text. There are two categories of text representation techniques: lexicon-based techniques and machine learning-based techniques. From research, both techniques have limitations. For instance, pre-trained word embeddings, such as Word2Vec, Glove, and bidirectional encoder representations from transformers (BERT), generate vectors by considering word distances, similarities, and occurrences ignoring other aspects such as word sentiment orientation. Aiming at such limitations, this paper presents a sentiment classification model (named LeBERT) combining sentiment lexicon, N-grams, BERT, and CNN. In the model, sentiment lexicon, N-grams, and BERT are used to vectorize words selected from a section of the input text. CNN is used as the deep neural network classifier for feature mapping and giving the output sentiment class. The proposed model is evaluated on three public datasets, namely, Amazon products’ reviews, Imbd movies’ reviews, and Yelp restaurants’ reviews datasets. Accuracy, precision, and F-measure are used as the model performance metrics. The experimental results indicate that the proposed LeBERT model outperforms the existing state-of-the-art models, with a F-measure score of 88.73% in binary sentiment classification.