15 research outputs found
An Intelligent Clustering Based Methodology for Confusable Diseases Diagnosis and Monitoring
No full textThe combination of non-specific clinical manifestations that characterize confusable tropical disease and the probable lack of expertise and experience among physicians exponentially increases the potential for misdiagnosis and subsequent increased morbidity and mortality rates resulting from these diseases. In this paper, an intelligent system driven by fuzzy clustering algorithm and Adaptive Neuro-Fuzzy Inference System for the investigation, diagnosis and management of similar and confusing symptoms of confusable diseases was developed. Data on patients diagnosed and confirmed by laboratory tests of viral hepatitis (H), malaria (M), typhoid fever (T) and urinary tract infection (U) were used for training, testing and validation of the system. The system assigns patients with severity levels in all the clusters. Results on clusters validity are satisfactory. Overlapping symptoms analysis shows that symptoms of both H and T have highest degree of overlapping while symptoms common to M and U yielded the least impact. Symptoms common to M, H and T only, have equal impact with that of M, T and U only. The symptoms that are common to all the four diseases under study yielded a 12.8% contribution to the degree of severity of each of the CTD diseases. The system compares favorably with diagnosis arrived at by experienced physicians and also provides patients’ level of severity in each confusable disease and the degree of confusability of any two or moreconfusable diseases.Keywords: Confusable diseases; viral hepatitis; malaria; typhoid fever; urinary tract infection; Clustering, ANFI
Comparative Analysis of Neural Network Models for Petroleum Products Pipeline Monitoring
No full textIn recent years, Neural Network (NN) has gained popularity in proffering solution to complex nonlinear problems. Monitoring of variations in Petroleum Products Pipeline (PPP) attributes (flow rate, pressure, temperature, viscosity, density, inlet and outlet volume) which changes with time is complex due to existence of non linear interaction amongst the attributes. The existing works on PPP monitoring are limited by lack of capabilities for pattern recognition and learning from previous data. In this paper, NN models with pattern recognition and learning capabilities are compared with a view of selecting the best model for monitoring PPP. Data was collected from Pipelines and Products Marketing Company (PPMC), Port Harcourt, Nigeria. The data was used for NN training, validation and testing with different NN models such as Multilayer Perceptron (MLP), Radial Basis Function (RBF), Generalized Feed Forward (GFF), Support Vector Machine (SVM), Time Delay Network (TDN) and Recurrent Neural Network (RNN). Neuro Solutions 6.0 was used as the front-end-engine for NN training, validation and testing while My Structured Query Language (MySQL) database served as the back-end-engine. Performance of NN models was measured using Mean Squared Error (MSE), Mean Absolute Error (MAE), Correlation Coefficient (r), Akaike Information Criteria (AIC) and Minimum Descriptive Length (MDL). MLP with one hidden layer and three processing elements performed better than other NN models in terms of MSE, MAE, AIC, MDL and r values between the computed and the desired output
Adaptive Cooperative Learning Methodology for Oil Spillage Pattern Clustering and Prediction
No full textThe serious environmental, economic and social consequences of oil spillages could devastate any nation of the world. Notable aftermath of this effect include loss of (or serious threat to) lives, huge financial losses, and colossal damage to the ecosystem. Hence, understanding the pattern and making precise predictions in real time is required (as opposed to existing rough and discrete prediction) to give decision makers a more realistic picture of environment. This paper seeks to address this problem by exploiting oil spillage features with sets of collected data of oil spillage scenarios. The proposed system integrates three state-of-the-art tools: self organizing maps, (SOM), ensembles of deep neural network (k-DNN) and adaptive neuro-fuzzy inference system (ANFIS). It begins with unsupervised learning using SOM, where four natural clusters were discovered and used in making the data suitable for classification and prediction (supervised learning) by ensembles of k-DNN and ANFIS. Results obtained showed the significant classification and prediction improvements, which is largely attributed to the hybrid learning approach, ensemble learning and cognitive reasoning capabilities. However, optimization of k-DNN structure and weights would be needed for speed enhancement. The system would provide a means of understanding the nature, type and severity of oil spillages thereby facilitating a rapid response to impending oils spillages. Keywords: SOM, ANFIS, Fuzzy Logic, Neural Network, Oil Spillage, Ensemble Learnin
A Hybrid Framework for Fire Outbreak Detection Based on Interval Type-2 Fuzzy Logic and Flower Pollination Algorithm
No full textRecognizing facial emotions for educational learning settings
No full textEducational learning settings exploit cognitive factors as ultimate feedback to enhance personalization in teaching and learning. But besides cognition, the emotions of the learner which reflect the affective learning dimension also play an important role in the learning process. The emotions can be recognized by tracking explicit behaviors of the learner like facial or vocal expressions. Despite reasonable efforts to recognize emotions, the research community is currently constraints by two issues, namely: i) the lack of efficient feature descriptors to accurately represent and prospectively recognize (detecting) the emotions of the learner; ii) lack of contextual datasets to benchmark performances of emotion recognizers in the learning-specific scenarios, resulting in poor generalizations. This paper presents a facial emotion recognition technique (FERT). The FERT is realized through results of preliminary analysis across various facial feature descriptors. Emotions are classified using the multiple kernel learning (MKL) method which reportedly possesses good merits. A contextually relevant simulated learning emotion (SLE) dataset is introduced to validate the FERT scheme. Recognition performance of the FERT scheme generalizes to 90.3% on the SLE dataset. On more popular but noncontextually datasets, the scheme achieved 90.0% and 82.8% respectively extended Cohn Kanade (CK+) and acted facial expressions in the wild (AFEW) datasets. A test for the null hypothesis that there is no significant difference in the performances accuracies of the descriptors rather proved otherwise (χ2=14.619,df=5,p=0.01212) for a model considered at a 95% confidence interval
Unsupervised Learning Framework for Customer Requisition and Behavioral Pattern Classification
No full textMaintaining healthy organization-customers relationship has positive influence on customers’ behavioral tendencies as regards preference to products and services, buying behavior, loyalty, satisfaction, and so on. To achieve this, an in-depth analysis of customers’ characteristics and purchasing behavioral trend is required. This paper proposes a hybrid unsupervised learning framework consisting of k-means algorithm and self-organizing maps (SOMs) for customer segmentation and behavior analysis. K-means algorithm was used to partition the entire input space of customers’ transaction dataset into 3 and 4 disjoint segments based on customers’ frequency (F) and monetary value (MV). SOM provided visualization of the underlying clusters and discovered customers’ relationships in the dataset. Interaction of F and MV clusters resulted in 12 sub-clusters. An in-depth analysis of each sub-cluster was also performed and appropriate customer relationship management (CRM) strategies established for each sub-cluster. Discovered knowledge will guide effective allocation of resources to each customer cluster and other organizational decision support functions much required by CRM systems.</jats:p
