Journal of Information Technology Management
Volume:15 Issue: 4, Autumn 2023

The realm of human-computer interaction delves deep into understanding how individuals acquire knowledge and integrate technology into their everyday lives. Among the various methods for measuring brain signals, electroencephalography (EEG) stands out for its non-invasive, portable, affordable, and highly time-sensitive capabilities. Some researchers have revealed a consistent correlation between meditation practices and changes in the EEG frequency range, observed across a wide array of meditation techniques. Furthermore, the availability of EEG datasets has facilitated research in this field. This study explores the effectiveness of the One-Dimensional Convolutional Neural Network (CNN-1D) based novel classification method, which impressively achieved an 62% training accuracy, showcasing the robustness of these models in meditation classification tasks. The proposed methodology unveiling a novel method to differentiate neural oscillations in 4 types of meditators and control. This approach analyzes an EEG dataset of highly experienced meditators practicing Vipassana (VIP), Isha Shoonya (SYN), Himalayan Yoga (HYT), and untrained control subjects (CTR) by employing chi-square, CNN, hyperparameter models for data analysis, The outcomes indicate that different meditation types exhibit distinct cognitive features, enabling effective differentiation and classification.

Co-Saliency object detection is the process of identifying common and repetitive objects from the group of images. Earlier studies have looked over several state-of-art deep neural network methodologies for co-saliency detection approach. The Deep CNN approaches rely heavily on co-saliency detection due to their potent feature extraction capabilities both deep and wide. This article assess the performance of several state-of-art deep learning model (VGG19, Inceptionv3, modifiedResNet, MobileNetV2 and PoolNet) for the purpose of co-saliency detection among images from benchmark datasets. All the models were trained on   70% part of the dataset and remaining were used for testing purpose. Experimental results show that modified ResNetmodel outperforms getting 96.53% accuracy as compared to other state-of-the-art deep neural network models.

Diseases like diabetes are chronic and require long-term management. Inadequate production of insulin results in high blood sugar levels. Such diseases lead to serious health issues such as heart ailments, blood vessel complaints, eye ailments, kidney function disorders, and nerve ailments. Hence, accurate assessment and management of risk factors are crucial for the onset of diabetes. Our proposed approach combines fuzzy logic & machine learning algorithms for diabetes risk prediction. Three machine learning models were trained to classify patients into two categories of diabetes (Type-I and Type-II) based on their clinical dataset collected from Katihar Medical College & Hospital and Suvadhan Lab. The polynomial regression algorithm achieved a score of 0.947, while the support vector regression algorithm with the rbf kernel achieved a score of 0.954, with a linear kernel achieved a score of 0.73. Our proposed approach performed well with respect to the conventional approaches with improved accuracy by identifying the patients at diabetes risk. In future work, we further analyze the relationship between other ignored factors which contribute to diabetes risk.

Internet-based social channels have turned into an important information repository for many people to get an idea about current trends and events happening around the world. As a result of Abundance of raw information on these social media platforms, it has become a crucial platform for businesses and individuals to make decisions based on social media analytics. The ever-expanding volume of online data available on the global network necessitates the use of specialized techniques and methods to effectively analyse and utilize this vast amount of information. This study's objective is to comprehend the textual information at the Lexical and Semantic level and to extract sentiments from this information in the most accurate way possible. To achieve this, the paper proposes to cluster semantically related words by evaluating their lexical similarity with respect to feature and sequence vectors. The proposed method utilizes Natural Language Processing, semantic and lexical clustering and hybrid C4.5 algorithm to extract six subcategories of emotions over three classes of sentiments based on word-based analysis of text. The proposed approach has yielded superior results with seven existing approaches in terms of parametric values, with an accuracy of 0.96, precision of 0.92, sensitivity of 0.94, and an f1-score of 0.92.

Numerous face recognition systems employ deep learning techniques to identify individuals in public areas such as shopping malls, airports, and other high-security zones. However, adversarial attacks are susceptible to deep learning-based systems. The adversarial attacks are intentionally generated by the attacker to mislead the systems. These attacks are imperceptible to the human eye. In this paper, we proposed a feature-based masking iterative method (F-MIM) to generate the adversarial images. In this method, we utilize the features of the face to misclassify the models. The proposed approach is based on a black-box attack technique where the attacker does not have the information related to target models. In this black box attack strategy, the face landmark points are modified using the binary masking technique. In the proposed method, we have used the momentum iterative method to increase the transferability of existing attacks. The proposed method is generated using the ArcFace face recognition model that is trained on the Labeled Face in the Wild (LFW) dataset and evaluated the performance of different face recognition models namely ArcFace, MobileFace, MobileNet, CosFace and SphereFace under the dodging and impersonate attack. The F-MIM attack is outperformed in comparison to the existing attacks based on Attack Success Rate evaluation metrics and further improves the transferability.

Due to obstruction in photosynthesis, the leaves of the plants get affected by the disease. Powdery mildew is the main disease in cucumber plants which generally occurs in the middle and late stages. Cucumber plant leaves are affected by various diseases, such as powdery mildew, downy mildew and Alternaria leaf spot, which ultimately affect the photosynthesis process; that’s why it is necessary to detect diseases at the right time to prevent the loss of plants. This paper aims to identify and classify diseases of cucumber leaves at the right time using a deep convolutional neural network (DCNN). In this work, the Deep-CNN model based on disease classification is used to enhance the performance of the ResNet50 model. The proposed model generates the most accurate results for cucumber disease detection using data enhancement based on a different data set. The data augmentation method plays an important role in enhancing the characteristics of cucumber leaves. Due to the requirements of the large number of parameters and the expensive computations required to modify standard CNNs, the pytorch library was used in this work which provides a wide range of deep learning algorithms. To assess the model accuracy large quantity of four types of healthy and diseased leaves and specific parameters such as batch size and epochs were compared with various machine learning algorithms such as support vector machine method, self-organizing map, convolutional neural network and proposed method in which the proposed DCNN model gave better results.

Net Asset Value (NAV) has long been a key performance metric for mutual fund investors. Due to the considerable fluctuation in the NAV value, it is risky for investors to make investment decisions. As a result, accurate and reliable NAV forecasts can help investors make better decisions and profit. In this research, we have analysed and compared the NAV prediction performance of our proposed deep learning models, such as N-BEATS and NBSL, with the FLANN model in both univariate and multivariate settings for five Indian mutual funds for forecast periods of 15, 20, 45, 63, 126, and 252 days using RMSE, MAPE, and R2 as evaluation metrics. A large forecast horizon was chosen to assess the model's consistency, reliability, and accuracy. The result reveals that the N-BEATS model outperforms the FLANN and NBSL models in the univariate setting for all datasets and all prediction horizons. In a multivariate setting, the outcome demonstrates that the N-BEATS model outperforms the FLANN model across all datasets and prediction horizons. The result also shows that, as the number of forecast days grew, our suggested models, notably N-BEATS, maintained consistency and attained the highest R2 value throughout the longest forecast duration.

Microfinance institutions in India provide a set of financial services to the economically weaker sections. Recently, a large number of microfinance institutions have emerged in India and they have favorable impact for poverty reduction. The impact of these institutions on entrepreneurship and society, needs to be explored in greater depth. The objective of this study is to apply machine learning techniques to explore this impact. The research uses a MIX dataset for three successive years, namely 2017, 2018, and 2019. This dataset comprises eight variables centered on gross loan portfolio. Principal Component Analysis (PCM) has been applied on the sample dataset for dimensionality reduction, resulting in two main components and each component consist of fraction from eight variables. Then, the sample dataset has been labelled with the help of clustering using K-means clustering technique. Further, classification models based on K-Nearest Neighbors (KNN) algorithm and Support Vector Machine (SVM) are applied to predict the appropriate category of entrepreneurship. The experiment result shows that the machine learning techniques have been found effective and useful tools for estimating the impact of microfinance on entrepreneurship in India.