Cryptocurrency is a popular digital currency due to its security and peer-to-peer transferability. Predicting cryptocurrency prices is crucial for investors and traders to make informed decisions on buying, selling, or holding cryptocurrencies based on their expected value, potential risks, and returns. This study aims to identify the optimal model for predicting the prices of cryptocurrencies, such as Bitcoin (BTC) and Ethereum (ETH), using Deephaven for Data curation. The study involves extracting data from both cryptocurrencies by Deephaven and selecting the most correlating parameters through time lag adjustment. We use correlating cryptocurrency data to train models, such as Artificial Neural Networks (ANN), Long-Short Term Memories (LSTM), and Gated Recurrent Units (GRU). Where the trial-and-error technique was applied for selecting optimized hyper-parameters for each model. The models are then evaluated by statistical evaluators, such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE), separately for training and testing datasets. For Bitcoin, the results showed that the LSTM model outperform ANN and GRU models in both training and testing data with MAE, RMSE, and MAPE average values of 0.079, 1.16, and 0.0006, respectively. While for Ethereum, the results also revealed that LSTM model performance is superior with MAE, RMSE, and MAPE average values of 0.0025, 0.124 and 0.0002, respectively. While GRU (MAE 0.012, RMSE 0.117, MAPE 0.002) performs robustly against ANN (MAE 0.035, RMSE 0.149, MAPE 0.003) model.

Internet of Things (IoT) provides a framework for interconnecting, controlling and monitoring of real-world objects and smart-devices through sensors. The sensors need energy sources to continuously operate. Since the design and the size constraints may impede a sensor node to have a distinct power source, the Energy Harvesting (EH) techniques emerge to cope with the energy constraint problem. Massive IoT networks with the Radio Frequency (RF) EH capability requires new accessing methods to reduce or mitigate interference and collisions. In this paper, we propose a group-based Medium Access Control (MAC) protocol, where a group head coordinates the corresponding nodes. The group heads working as Hybrid Access Points (HAPs) provide nodes with power and relay their data to the base station (BS). The HAPs use a distributed coordination function (DCF) to access the channel, along with in-group coordinating mechanism. In a group, nodes reduce collisions by using a frame slotted ALOHA to access the medium, with the addition of the EH mechanism. Nodes utilize the harvest-transmit-and-sleep protocol to harvest energy, to transmit information utilizing the harvested energy and to sleep when channel is inactive. We study channel models for different communication environments influencing the usability of RF energy harvesting and examine their impact on the network performance. We aim to mitigate inter-group and intra-group collisions and to provide an energy harvesting and data relaying mechanism in a distributed manner. Performance evaluations are carried out through extensive simulations and the model and analysis are verified. It is shown that the analysis results match well with those of simulations. The proposed distributed group-coordination mechanism achieves significant network performance gain over the conventional schemes. The effectiveness of the proposed scheme is shown by the network throughput, and the energy efficiency.

Decentralized cryptocurrencies like Bitcoin are digital assets with a price volatility nature, that allow for blockchain-based, peer-to-peer monetary transactions. Due to the price volatility problem with decentralized cryptocurrencies, research into the underlying pricing mechanism is required. Additionally, the behavior of Bitcoin prices is non-stationary, meaning that the statistical distribution of data varies over time. The proposed framework demonstrates the use of sophisticated machine learning models in predicting the short and medium-term trends and actual values of Bitcoin prices. This research goes beyond previous work that has only looked at machine learning-based categorization for a single day by instead using such models to forecast price changes seven, thirty, and ninety days into the future. The generated models are useful and work admirably, with the classification models reaching a maximum accuracy enhancement up to 31.48% for a 90-day prediction and a 11.76% F1-score for a forecast extending to the thirtieth day. In the case of regression the margin of error shifts from the one-time horizon for price projections to the next. A significant notable downfall occurs in different error metrics. These findings suggest that the models given here outperform those already found in the literature.

In this paper, we consider a relay-enabled wireless-powered communication network (WPCN), consisting of a relay-hybrid access point (RHAP), harvesting and transmitting nodes and a Base Station (BS). Contrary to conventional networks, the nodes rely on the energy signal of RHAP to harvest energy, and they transmit information to the BS via the RHAP. The nodes work on the principle of harvest-transmit-to-cooperate (HTTC) protocol, in which a node harvests energy from a power source in the downlink and transfers information to a relay in the uplink to convey data to the BS. In this paper, we propose a new distributed Medium Access Control (MAC) mechanism in a relay WPCN, based on carrier sense multiple access with collision avoidance (CSMA/CA), a binary exponential random backoff mechanism. In the proposed scheme, the nodes and the RHAP contend to access the channel with different contention window sizes. Furthermore, the RHAP is given more preference to access the channel by giving a smaller contention window size compared to the nodes.We propose a Markov chain model for the proposed- MAC and a Markov chain model for the energy states of nodes. It models the behavior of changing states between energy harvesting and data transmission. The analysis and simulations show the effectiveness of the proposed MAC in terms of throughput and energy efficiency.

Speech is a powerful means of expressing thoughts, emotions, and perspectives. However, accurately determining the emotions conveyed through speech remains a challenging task. Existing manual methods for analyzing speech to recognize emotions are prone to errors, limiting our understanding and response to individuals’ emotional states. To address diverse accents, an automated system capable of real-time emotion prediction from human speech is needed. This paper introduces a speech emotion recognition (SER) system that leverages supervised learning techniques to tackle cross-accent diversity. Distinctively, the system extracts a comprehensive set of nine speech features—Zero Crossing Rate, Mel Spectrum, Pitch, Root Mean Square values, Mel Frequency Cepstral Coefficients, chroma-stft, and three spectral features (Centroid, Contrast, and Roll-off) for refined speech signal processing and recognition. Seven machine learning models are employed, encompassing Random Forest, Logistic Regression, Decision Tree, Support Vector Machines, Gaussian Naive Bayes, K-Nearest Neighbors, ensemble learning, and four individual, hybrid deep learning models including Long short-term memory (LSTM) and 1-Dimensional Convolutional Neural Network (1D-CNN) with stratified cross-validation. Audio samples from diverse English regions are combined to train the models. The performance evaluation results of conventional machine learning and deep learning models indicate that the Random Forest-based feature selection model achieves the highest accuracy of up to 76% among the conventional machine learning models. Simultaneously, the 1D-CNN model with stratified cross-validation reaches up to 99% accuracy. The proposed framework enhances the cross-accent emotion recognition accuracy up to 86.3%, 89.87%, 90.27%, and 84.96% by margins of 14.71%, 10.15%, 9.6%, and 16.52% respectively.

Gastrointestinal (GI) disorders are common and often debilitating health issues that affect a significant portion of the population. Recent advancements in artificial intelligence, particularly computer vision algorithms, have shown great potential in detecting and classifying medical images. These algorithms utilize deep convolutional neural network architectures to learn complex spatial features in images and make predictions for similar unseen images. The proposed study aims to assist gastroenterologists in making more efficient and accurate diagnoses of GI patients by utilizing its two-phase transfer learning framework to identify GI diseases from endoscopic images. Three pre-trained image classification models, namely Xception, InceptionResNetV2, and VGG16, are fine-tuned on publicly available datasets of annotated endoscopic images of the GI tract. Additionally, two custom convolutional neural networks are constructed and fully trained for comparative analysis of their performance. Four different classification tasks are examined based on the endoscopic image categories. The proposed architecture employing InceptionResNetV2 achieves the most consistent and generalized performance across most classification tasks, yielding accuracy scores of 85.7% for general classification of GI tract (eight-category classification), 97.6% for three-diseases classification, 99.5% for polyp identification (binary classification), and 74.2% for binary classification of esophagitis severity on unseen endoscopic images. The results indicate the effectiveness of the two-phase transfer learning framework for clinical use to enhance the identification of GI diseases, aiding in their early diagnosis and treatment.