International Journal of Advances in Intelligent Informatics
Not a member yet
235 research outputs found
Sort by
An efficient activity recognition for homecare robots from multi-modal communication dataset
Full text linkHuman environments are designed and managed by humans for humans. Thus, adding robots to interact with humans and perform specific tasks appropriately is an essential topic in robotics research. In recent decades, object recognition, human skeletal, and face recognition frameworks have been implemented to support the tasks of robots. However, recognition of activities and interactions between humans and surrounding objects is an ongoing and more challenging problem. Therefore, this study proposed a graph neural network (GNN) approach to directly recognize human activity at home using vision and speech teaching data. Focus was given to the problem of classifying three activities, namely, eating, working, and reading, where these activities were conducted in the same environment. From the experiments, observations, and analyses, this proved to be quite a challenging problem to solve using only traditional convolutional neural networks (CNN) and video datasets. In the proposed method, an activity classification was learned from a 3D detected object corresponding to the human position. Next, human utterances were used to label the activity from the collected human and object 3D positions. The experiment, involving data collection and learning, was demonstrated by using human-robot communication. It was shown that the proposed method had the shortest training time of 100.346 seconds with 6000 positions from the dataset and was able to recognize the three activities more accurately than the deep layer aggregation (DLA) and X3D networks with video datasets
Contrast enhancement for improved blood vessels retinal segmentation using top-hat transformation and otsu thresholding
Full text linkDiabetic Retinopathy is a effect of diabetes. It results abnormalities in the retinal blood vessels. The abnormalities can cause blurry vision and blindness. Automatic retinal blood vessels segmentation on retinal image can detect abnormalities in these blood vessels, actually resulting in faster and more accurate segmentation results. The paper proposed an automatic blood vessel segmentation method that combined Otsu Thresholding with image enhancement techniques. In image enhancement, it combined CLAHE with Top-hat transformation to improve image quality. The study used DRIVE dataset that provided retinal image data. The image data in dataset was generated by the fundus camera. The CLAHE and Top-hat transformation methods were applied to rise the contrast and reduce noise on the image. The images that had good quality could help the segmentation process to find blood vessels in retinal images appropriately by a computer. It improved the performance of the segmentation method for detecting blood vessels in retinal image. Otsu Thresholding was used to segment blood vessel pixels and other pixels as background by local threshold. To evaluation performance of the proposed method, the study has been measured accuracy, sensitivity, and specificity. The DRIVE dataset's study results showed that the averages of accuracy, sensitivity, and specificity values were 94.7%, 72.28%, and 96.87%, respectively. It indicated that the proposed method was successful and well to work on blood vessels segmentation retinal images especially for thick blood vessels
Exploration of hybrid deep learning algorithms for covid-19 mrna vaccine degradation prediction system
Full text linkCoronavirus causes a global pandemic that has adversely affected public health, the economy, including every life aspect. To manage the spread, innumerable measurements are gathered. Administering vaccines is considered to be among the precautionary steps under the blueprint. Among all vaccines, the messenger ribonucleic acid (mRNA) vaccines provide notable effectiveness with minimal side effects. However, it is easily degraded and limits its application. Therefore, considering the cruciality of predicting the degradation rate of the mRNA vaccine, this prediction study is proposed. In addition, this study compared the hybridizing sequence of the hybrid model to identify its influence on prediction performance. Five models are created for exploration and prediction on the COVID-19 mRNA vaccine dataset provided by Stanford University and made accessible on the Kaggle community platform employing the two deep learning algorithms, Long Short-Term Memory (LSTM) as well as Gated Recurrent Unit (GRU). The Mean Columnwise Root Mean Square Error (MCRMSE) performance metric was utilized to assess each model’s performance. Results demonstrated that both GRU and LSTM are befitting for predicting the degradation rate of COVID-19 mRNA vaccines. Moreover, performance improvement could be achieved by performing the hybridization approach. Among Hybrid_1, Hybrid_2, and Hybrid_3, when trained with Set_1 augmented data, Hybrid_3 with the lowest training error (0.1257) and validation error (0.1324) surpassed the other two models; the same for model training with Set_2 augmented data, scoring 0.0164 and 0.0175 MCRMSE for training error and validation error, respectively. The variance in results obtained by hybrid models from experimenting claimed hybridizing sequence of algorithms in hybrid modeling should be concerned
Incremental multiclass open-set audio recognition
Full text linkIncremental learning aims to learn new classes if they emerge while maintaining the performance for previously known classes. It acquires useful information from incoming data to update the existing models. Open-set recognition, however, requires the ability to recognize examples from known classes and reject examples from new/unknown classes. There are two main challenges in this matter. First, new class discovery: the algorithm needs to not only recognize known classes but it must also detect unknown classes. Second, model extension: after the new classes are identified, the model needs to be updated. Focusing on this matter, we introduce incremental open-set multiclass support vector machine algorithms that can classify examples from seen/unseen classes, using incremental learning to increase the current model with new classes without entirely retraining the system. Comprehensive evaluations are carried out on both open set recognition and incremental learning. For open-set recognition, we adopt the openness test that examines the effectiveness of a varying number of known/unknown labels. For incremental learning, we adapt the model to detect a single novel class in each incremental phase and update the model with unknown classes. Experimental results show promising performance for the proposed methods, compared with some representative previous methods
Land cover classification based optical satellite images using machine learning algorithms
Full text linkThis article aims to apply machine learning algorithms to the supervised classification of optical satellite images. Indeed, the latter is efficient in the study of land use. Despite the performance of machine learning in satellite image processing, this can change but depends on the nature of the satellite images used. Moreover, when we use the satellite, then the reliability of one classifier can be different from the others. In this paper, we examined the performance of DT, SVM, KNN, ANN, and RF. Analysis factors were used to investigate further their importance for Sentinel 2, Landsat 8, Terra Modis, and Spot 5 images. The results show that the KNN showed the most interesting accuracy during the analysis of medium and low-resolution images with spectral bands lower or equal to 4, with a higher accuracy of about 93%. The RF completely dominated the other analysis cases, where the higher accuracy was about 94%. The classification accuracy is more reliable with high-resolution images than with the other resolution categories. However, the processing times of high-resolution images are much higher. Moreover, higher accuracy was often achieved with more expensive processing times. Besides, almost all machine learning algorithms suffered from the Hugs phenomenon during the analyses. So, before the classification with machine learning, some preprocessing is needed
Aspect-based sentiment analysis for hotel reviews using an improved model of long short-term memory
Full text linkAdvances in information technology have given rise to online hotel reservation options. The user review feature is an important factor during the online booking of hotels. Generally, most online hotel booking service providers provide review and rating features for assessing hotels. However, not all service providers provide rating features or recap reviews for every aspect of the hotel services offered. Therefore, we propose a method to summarise reviews based on multiple aspects, including food, room, service, and location. This method uses long short-term memory (LSTM), together with hidden layers and automation of the optimal number of hidden neurons. The F1-measure value of 75.28% for the best model was based on the fact that (i) the size of the first hidden layer is 1,200 neurons with the tanh activation function, and (ii) the size of the second hidden layer is 600 neurons with the ReLU activation function. The proposed model outperforms the baseline model (also known as standard LSTM) by 10.16%. It is anticipated that the model developed through this study can be accessed by users of online hotel booking services to acquire a review recap on more specific aspects of services offered by hotel
An extended approach of weight collective influence graph for detection influence actor
Full text linkOver the last decade, numerous methods have been developed to detect the influential actors of hate speech in social networks, one of which is the Collective Influence (CI) method. However, this method is associated with unweighted datasets, which makes it inappropriate for social media, significantly using weight datasets. This study proposes a new CI method called the Weighted Collective Influence Graph (WCIG), which uses the weights and neighbor values to detect the influence of hate speech. A total of 49, 992 Indonesian tweets were and extracted from Indonesian Twitter accounts, from January 01 to January 22, 2021. The data collected are also used to compare the results of the proposed WCIG method to determine the influential actors in the dissemination of information. The experiment was carried out two times using parameters ∂=2 and ∂=4. The results showed that the usernames bernacleboy and zack_rockstar are influential actors in the dataset. Furthermore, the time needed to process WCIG calculations on HPC is 34-75 hours because the larger the parameter used, the greater the processing time
Optimizing complexity weight parameter of use case points estimation using particle swarm optimization
Full text linkAmong algorithmic-based frameworks for software development effort estimation, Use Case Points I s one of the most used. Use Case Points is a well-known estimation framework designed mainly for object-oriented projects. Use Case Points uses the use case complexity weight as its essential parameter. The parameter is calculated with the number of actors and transactions of the use case. Nevertheless, use case complexity weight is discontinuous, which can sometimes result in inaccurate measurements and abrupt classification of the use case. The objective of this work is to investigate the potential of integrating particle swarm optimization (PSO) with the Use Case Points framework. The optimizer algorithm is utilized to optimize the modified use case complexity weight parameter. We designed and conducted an experiment based on real-life data set from three software houses. The proposed model’s accuracy and performance evaluation metric is compared with other published results, which are standardized accuracy, effect size, mean balanced residual error, mean inverted balanced residual error, and mean absolute error. Moreover, the existing models as the benchmark are polynomial regression, multiple linear regression, weighted case-based reasoning with (PSO), fuzzy use case points, and standard Use Case Points. Experimental results show that the proposed model generates the best value of standardized accuracy of 99.27% and an effect size of 1.15 over the benchmark models. The results of our study are promising for researchers and practitioners because the proposed model is actually estimating, not guessing, and generating meaningful estimation with statistically and practically significant
An Improved particle swarm optimization based on lévy flight and simulated annealing for high dimensional optimization problem
Full text linkParticle swarm optimization (PSO) is a simple metaheuristic method to implement with robust performance. PSO is regarded as one of the numerous researchers' most well-studied algorithms. However, two of its most fundamental problems remain unresolved. PSO converges onto the local optimum for high-dimensional optimization problems, and it has slow convergence speeds. This paper introduces a new variant of a particle swarm optimization algorithm utilizing Lévy flight-McCulloch, and fast simulated annealing (PSOLFS). The proposed algorithm uses two strategies to address high-dimensional problems: hybrid PSO to define the global search area and fast simulated annealing to refine the visited search region. In this paper, PSOLFS is designed based on a balance between exploration and exploitation. We evaluated the algorithm on 16 benchmark functions for 500 and 1,000 dimension experiments. On 500 dimensions, the algorithm obtains the optimal value on 14 out of 16 functions. On 1,000 dimensions, the algorithm obtains the optimal value on eight benchmark functions and is close to optimal on four others. We also compared PSOLFS with another five PSO variants regarding convergence accuracy and speed. The results demonstrated higher accuracy and faster convergence speed than other PSO variants. Moreover, the results of the Wilcoxon test show a significant difference between PSOLFS and the other PSO variants. Our experiments' findings show that the proposed method enhances the standard PSO by avoiding the local optimum and improving the convergence speed
Semi-supervised learning for sentiment classification with ensemble multi-classifier approach
Full text linkSupervised sentiment analysis ideally uses a fully labeled data set for modeling. However, this ideal condition requires a struggle in the label annotation process. Semi-supervised learning (SSL) has emerged as a promising method to avoid time-consuming and expensive data labeling without reducing model performance. However, the research on SSL is still limited and its performance needs to be improved. Thus, this study aims to create a new SSL-Model for sentiment analysis. The Ensemble Classifier SSL model for sentiment classification is introduced. The research went through pre-processing, vectorization, and feature extraction using TF-IDF and n-grams. Support Vector Machine (SVM) or Random Forest for tokenization was used to separate unigram, bigram, and trigram in model generation. Then, the outputs of these models were combined using stacking ensemble approach. Accuracy and F1-score were used for the evaluation. IMDB datasets and US Airlines were used to test the new SSL models. The conclusion is that the sentiment annotation accuracy is highly dependent on the suitability of the dataset with the machine learning algorithm. In IMDB dataset, which consists of two classes, it is better to use SVM. In the US Airlines consisting of three classes, SVM is better at improving the model performance against the baseline, but RF is better at achieving the baseline performance even though it fails to maintain the model performance