Encoding Technique Research Articles

Application of machine learning for analyzing cancer patient data and predicting survival

Cancer is a deadly disease, and a leading cause of death globally. Thus, the prediction of the possibility of survival of cancer patients, at an early stage of treatment, will be beneficial for both the doctors and the patients. This study has attempted to predict the survival status of cancer patients, by employing two well-known Machine Learning algorithms viz., Logistic Regression and Support Vector Machine, and utilizing a dataset of Kaggle. Before using the Machine Learning models, suitable encoding and scaling techniques have been applied on the data. However, neither of the Machine Learning algorithms has performed satisfactorily (accuracy of prediction for Logistic Regression: 51.6%, and that for Support Vector Machine: 52.2%), and the actual reason for this poor performance seems to be the low quality and/or the insufficiency of the data used.

A temporal knowledge graph reasoning model based on recurrent encoding and contrastive learning.

Temporal knowledge graphs (TKGs) are critical tools for capturing the dynamic nature of facts that evolve over time, making them highly valuable in a broad spectrum of intelligent applications. In the domain of temporal knowledge graph extrapolation reasoning, the prediction of future occurrences is of great significance and presents considerable obstacles. While current models consider the fact changes over time and recognize that historical facts may recur, they often overlook the influence of past events on future predictions. Motivated by these considerations, this work introduces a novel temporal knowledge graph reasoning model, named Temporal Reasoning with Recurrent Encoding and Contrastive Learning (TRCL), which integrates recurrent encoding and contrastive learning techniques. The proposed model has the ability to capture the evolution of historical facts, generating representations of entities and relationships through recurrent encoding. Additionally, TRCL incorporates a global historical matrix to account for repeated historical occurrences and employs contrastive learning to alleviate the interference of historical facts in predicting future events. The TKG reasoning outcomes are subsequently derived through a time decoder. A quantity of experiments conducted on four benchmark datasets demonstrate the exceptional performance of the proposed TRCL model across a range of metrics, surpassing state-of-the-art TKG reasoning models. When compared to the strong baseline Time-Guided Recurrent Graph Network (TiRGN) model, the proposed TRCL achieves 1.03% improvements on ICEWS14 using mean reciprocal rank (MRR) evaluation metric. This innovative proposed method not only enhances the accuracy of TKG extrapolation, but also sets a new standard for robustness in dynamic knowledge graph applications, paving the way for future research and practical applications in predictive intelligence systems.

Artificial intelligence virtual assistants in primary eye care practice.

To propose a novel artificial intelligence (AI)-based virtual assistant trained on tabular clinical data that can provide decision-making support in primary eye care practice and optometry education programmes. Anonymised clinical data from 1125 complete optometric examinations (2250 eyes; 63% women, 37% men) were used to train different machine learning algorithm models to predict eye examination classification (refractive, binocular vision dysfunction, ocular disorder or any combination of these three options). After modelling, adjustment, mining and preprocessing (one-hot encoding and SMOTE techniques), 75 input (preliminary data, history, oculomotor test and ocular examinations) and three output (refractive, binocular vision status and eye disease) features were defined. The data were split into training (80%) and test (20%) sets. Five machine learning algorithms were trained, and the best algorithms were subjected to fivefold cross-validation. Model performance was evaluated for accuracy, precision, sensitivity, F1 score and specificity. The random forest algorithm was the best for classifying eye examination results with a performance >95.2% (based on 35 input features from preliminary data and history), to propose a subclassification of ocular disorders with a performance >98.1% (based on 65 features from preliminary data, history and ocular examinations) and to differentiate binocular vision dysfunctions with a performance >99.7% (based on 30 features from preliminary data and oculomotor tests). These models were integrated into a responsive web application, available in three languages, allowing intuitive access to the AI models via conventional clinical terms. An AI-based virtual assistant that performed well in predicting patient classification, eye disorders or binocular vision dysfunction has been developed with potential use in primary eye care practice and education programmes.

Convolutional auto-encoder and discrete wavelet transform for lossy region-based medical image compression

Medical imaging is a vital and ever-evolving discipline that plays a critical role in detecting, diagnosing, and planning surgeries for diseases. In medical imaging, image compression systems are employed to reduce storage and bandwidth requirements. This paper presents a region of interest (ROI) based image compression method for brain MRI images by employing a convolutional auto-encoder (CAE) and discrete wavelet transform (DWT) to achieve both efficient compression and high-quality reconstruction in terms of lossy compression. Our compression approach involves separating image regions, employing a lossy compression technique based on CAE and DWT, with a lossless compression technique based arithmetic encoding. To achieve optimal compression of non-ROI regions, the encoding network of the CAE generates a compacted feature map that preserves structural information, which aids the DWT based codec and the decoding network of the CAE in reconstructing the output. To maintain diagnostically significant information, the ROI portion of the image is losslessly compressed using arithmetic encoding technique. To assess the effectiveness of our proposed method, we compared our results to standard lossy compression algorithms and recent approaches. Our method achieved a peak signal-to-noise ratio (PSNR) of 37.91 dB and a mean structural similarity index measure (MS-SSIM) of 98.62% at a high compression ratio of 30.

Neural deflection field for sparse-view tomographic background oriented Schlieren

Three-dimensional (3D) density-varying turbulent flows are widely encountered in high-speed aerodynamics, combustion, and heterogeneous mixing processes. Multicamera-based tomographic background-oriented Schlieren (TBOS) has emerged as a powerful technique for revealing 3D flow density structures. However, dozens of cameras are typically required to obtain high-quality reconstructed density fields. Limited by the number of available optical windows and confined space in the harsh experimental environments, TBOS with only sparse views and limited viewing angles often becomes the necessary choice practically, rendering the inverse problem for TBOS reconstruction severely ill-posed and resulting in degraded tomography quality. In this study, we propose a novel TBOS reconstruction method, neural deflection field, utilizing an extremely light-weight deep neural networks to represent the density gradient fields without using any pretrained neural network models. Particularly, state-of-the-art positional encoding techniques and hierarchical sampling strategies are incorporated to capture the density structures of high spatial frequencies. Required background images for TBOS reconstructions are synthesized based on a high-fidelity nonlinear ray-tracing method with the ground truth flows from conducting large eddy simulations on premixed turbulent flames. Owing to these synthesized BOS images, the superiority of the proposed method is quantitatively verified compared to the classical TBOS reconstruction methods, and the specific contributions from the position encoding and the hierarchical sampling strategy are also elucidated.

IMPLEMENTATION OF MULTIPLE LINEAR REGRESSION ALGORITHM IN PREDICTING RED CHILI PRICES IN GARUT REGENCY

Vegetables, including red chili peppers, play an important role in food and economic balance. Significant price fluctuations and inflation are often problems for farmers and traders. Garut Regency, as the center of red chili production in West Java, faces similar challenges. This research aims to implement a Multiple Linear Regression algorithm to predict the price of red chili peppers in the Garut Regency, highlighting the novelty of using a combination of One Hot Encoding, Feature Engineering, Standard Scaler, and Hyperparameter Tuning techniques. The method used is CRISP-DM with 6 stages: Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. The data used is the price and production of red chili peppers per week in 2018-2023, with a total of 702 records. This research involved 8 trials with data transformation and normalization scenarios. The model evaluation used MSE, RMSE, MAPE, R-squared, and statistical hypothesis testing metrics. Results showed 5 significantly influential attributes: year, month, production, net harvested area, and productivity. The best model yielded MSE 202,134,650, RMSE 14,217, MAPE 29.16%, and R-squared 0.320. This approach is simpler yet effective and is able to provide fairly accurate predictions. This research is expected to contribute to providing predictive models that help farmers and traders anticipate price fluctuations, as well as provide insights for policymakers in price management.

Enhanced spectral reconstruction of ultrafast spatiotemporal encoded 2D NMR spectroscopy

BackgroundNuclear Magnetic Resonance (NMR) is extensively utilized in research as a non-invasive technique for investigating molecular structures and composite components. The spatiotemporal encoding (SPEN) technique effectively accelerates multi-dimensional NMR experiments. In ultrafast SPEN NMR, the acquired data are divided into odd and even segments corresponding to the positive and negative gradients during the decoding stage, respectively. However, the interlaced Fourier transform (FT) method used to reconstruct a full-width spectrum from these segments often suffers from severe noise contamination, necessitating the development of a more effective spectrum reconstruction method. ResultsIn this work, we analyze the noise amplification effect of the interlaced FT and find that the noise is most significant in two edge regions of the spectrum along the indirect dimension due to the relatively small time offset differences between odd and even segments in those regions. Consequently, we develop an iterative optimization method to obtain the full-width spectrum while mitigating the noise. The proposed method incorporates the odd and even data segments into an objective function with sparsity regularization to simplify the spectrum, which is then refined iteratively during the optimization. As a result, the reconstructed spectrum is significantly cleaner and maintains the full spectral width. Experimental results demonstrate a remarkable improvement in the readability and interpretability of SPEN data, evidenced by clearer signal peaks and reduced background noise. SignificanceThe proposed reconstruction method provides a reliable approach for processing SPEN 2D NMR data, effectively addressing the low sensitivity issue in the joint reconstruction on odd and even segments. Combining SPEN's ultrafast data acquisition with the proposed high-sensitivity spectrum reconstruction method enhances the utility of NMR for more accurate molecular structure analysis and component identification in composite samples, particularly promoting NMR research in rapid reaction systems.

Robust image encryption algorithm based on oscillated substitution and effective confusion module with novel chaining permutation and pixel mutation

To ensure secure network transmission of multimedia content, it is crucial to protect it from unauthorized access. Encryption techniques are de facto protection techniques against hacking attempts. Different methods, including mathematical encoding, neural networks, DNA encoding, and chaos-based techniques, have been developed for this purpose. However, these methods often overlook the impact of input variations, which can enhance encryption robustness by amplifying confusion. We propose an encryption technique that demonstrates how effectively leveraging input variations can enhance the complexity and security of encrypted information. We designed an oscillated substitution technique, a multi-state method that adjusts the substitution pattern based on changes in the key or input. Additionally, chaotic methods are employed to generate noise codes for bit mutation and chaining-based permutation. Experimental results indicate that our technique significantly improves security indicators and outperforms recent encryption methods.

Precise Orientation Estimation for Rotated Object Detection Based on a Unit Vector Coding Approach

Existing rotated object detection methods usually use angular parameters to represent the object orientation. However, due to the symmetry and periodicity of these angular parameters, a well-known boundary discontinuity problem often results. More specifically, when the object orientation angle approaches the periodic boundary, the predicted angle may change rapidly and adversely affect model training. To address this problem, this paper introduces a new method that can effectively solve the boundary discontinuity problem related to angle parameters in rotated object detection. Our approach involves a novel vector-based encoding and decoding technique for angular parameters, and a cosine distance loss function for angular accuracy evaluation. By utilizing the characteristics of unit vectors and cosine similarity functions, our method parameterizes the orientation angle as components of the unit vector during the encoding process and redefines the orientation angle prediction task as a vector prediction problem, effectively avoiding the boundary discontinuity problem. The proposed method achieved a mean average precision (mAP) of 87.48% and an average cosine similarity (CS) of 0.997 on the MVTec test set. It also achieved an mAP score of 90.54% on the HRSC2016 test set, which is better than several existing state-of-the-art methods and proves its accuracy and effectiveness.

Movie Recommendation System Based on Word Embedding

Abstract. Amidst the swift progression of Internet and big data technologies, recommendation systems have emerged as crucial conduits linking users to products across various digital platforms. This study delves into the deployment of neural collaborative filtering within the realm of movie recommendation systems, with the objective of constructing a system of high precision. Utilizing the MovieLens dataset, this investigation applies one-hot encoding and embedding techniques within the PyTorch-Lightning framework to effectively model user behaviors and predict cinematic preferences. The neural collaborative filtering (NCF) model leverages deep learning to extract latent features of users and items, exhibiting a notable enhancement in performance compared to traditional collaborative filtering approaches. Empirical results indicate that the NCF model attains an accuracy of 86% on the validation dataset, underscoring its efficacy in extensive recommendation scenarios.

LSTGCN: Inductive Spatial Temporal Imputation Using Long Short-Term Dependencies

Spatial temporal forecasting of urban sensors is essentially important for many urban systems, such as intelligent transportation and smart cities. However, due to the problem of hardware failure or network failure, there are some missing values or missing monitoring sensors that need to be interpolated. Recent research on deep learning has made substantial progress on imputation problem, especially temporal aspect (i.e., time series imputation), while little attention has been paid to spatial aspect (both dynamic and static) and long-term temporal dependencies. In this article, we proposed a spatial temporal imputation model, named Long Short-Term Graph Convolution Networks (LSTGCN), which includes gated temporal extraction (GTE) module, multi-head attention-based temporal capture (MHAT) module, long-term periodic temporal encoding (LPTE) module, and bidirectional spatial graph convolution (BSGC) module. The GTE adopts a gated mechanism to filter short-term temporal information, while the MHAT utilizes position encoding to enhance the difference of each timestamps, then use multi-head attention to capture short-term temporal dependency. The BSGC is adopted to handle with spatial relationships between sensor nodes. And we design a periodic encoding technique to process long-term temporal dependencies. The BSGC handles spatial relationships between sensor nodes, and a periodic encoding technique is used to process long-term temporal dependencies. Our experimental analysis includes completion and forecasting tasks, as well as transfer and ablation analyses. The results show that our proposed model outperforms state-of-the-art baselines on real-world datasets.

Cryptanalyzing an Image Encryption Scheme Using Synchronization of Memristor Chaotic Systems

This paper conducts a security analysis on an image encryption scheme that incorporates chaos synchronization and DNA encoding techniques. The scheme adheres to the typical Substitution–Permutation structure by introducing the DNA operation as a nonlinear component. Compared to the traditional chaotic ciphers, the designer claims that the scheme achieved better nonlinear properties through the DNA operation. However, the characteristics of the differential between some pairs of plaintexts and the corresponding ciphertexts can be exploited to break the permutation operation. Meanwhile, the substitution operation can also be cracked by checking the mapping between the intermediate variables inversely permuted from ciphertext and the corresponding plaintext. Given the weaknesses of the encryption scheme, we introduce effective strategies for the known-plaintext attack. Both rigorous theoretical analyses and detailed experimental results are provided to demonstrate the effectiveness of the attacks.

Coded speckle target: Design, analysis and applications

In this paper, a novel encoding feature based on speckle patterns. It comprises three components: a pre-positioning ring, localization speckle area, and encoding speckle area. The pre-positioning ring identifies potential regions of interest. Shape function of the localization speckle area can be estimated by formulating the elliptical equation of the pre-positioning ring. Next, digital image correlation method is employed to achieve precise positioning and refine the shape function. Through the high-precision shape function, the encoded speckle area can be determined. Encoding technique utilizes a reference encoding scheme, in which each coded value region either same or is inverse to the reference region. Zero-mean normalized cross correlation relative to the reference region can be used to obtain a highly robust decoding value. Experimental results demonstrate that, despite challenges such as noise, defocus, variations in lighting intensity, and lighting uneven, the coded speckle target maintains high robustness and precise positioning accuracy. Moreover, the structure from motion and displacement measurement experiments further verifies its high accuracy in practical applications.

A refined approach to early movie box office prediction leveraging ensemble learning and feature encoding

Predicting the revenue of a movie prior to its release presents a significant challenge. The ability to predict pre-release revenue enables movie production companies to devise effective marketing strategies and mitigate the risks associated with potential box office failures. The primary hurdles in this endeavor stem from managing the myriad factors influencing box office outcomes and accurately forecasting a movie's revenue before it becomes available to the public. To overcome these challenges, we introduce a sophisticated Early Movie Box Office Prediction Model that incorporates Ensemble Learning and Feature Encoding techniques. This model amalgamates multiple foundational models, utilizing regression and decision trees to forecast box office revenues. Our composite model demonstrates superior performance over its constituent models, achieving an impressive accuracy rate of 91.4%.

Plant Leaf Disease Detection and Classification Using Segmentation Encoder Techniques

Aims Agriculture is one of the fundamental elements of human civilization. Crops and plant leaves are susceptible to many illnesses when grown for agricultural purposes. There may be less possibility of further harm to the plants if the illnesses are identified and classified accurately and early on. Background Plant leaf diseases are typically predicted and classified by farmers tediously and inaccurately. Manual identification of diseases may take more time and may not accurately detect the disease. There could be a major drop in production if crop plants are destroyed due to slow detection and classification of plant illnesses. Radiologists used to segment leaf lesions manually, which takes a lot of time and work. Objective It is established that deep learning models are superior to human specialists in the diagnosis of lesions on plant leaves. Here, the “Deep Convolutional Neural Network (DCNN)” based encoder-decoder architecture is suggested for the semantic segmentation of leaf lesions. Methods A proposed semantic segmentation model is based on the Dense-Net encoder. The LinkNet-34 segmentation model performance is compared with two other models, SegNet and PSPNet. Additionally, the two encoders, ResNeXt and InceptionV3, have been compared to the performance of DenseNet-121, the encoder used in the LinkNet-34 model. After that, two different optimizers, such as Adam and Adamax, are used to optimize the proposed model. Results The DenseNet-121 encoder utilizing Adam optimizer has been outperformed by the LinkNet-34 model, with a dice coefficient of 95% and a Jaccard Index of 93.2% with a validation accuracy of 97.57%. Conclusion The detection and classification of leaf disease with deep learning models gives better results in comparison with other models.

A Study on the Man-Hour Prediction in Structural Steel Fabrication

Longitudinal cutting is the most common process in steel structure manufacturing, and the man-hours of the process provide an important basis for enterprises to generate production schedules. However, currently, the man-hours in factories are mainly estimated by experts, and the accuracy of this method is relatively low. In this study, we propose a system that predicts man-hours with history data in the manufacturing process and that can be applied in practical structural steel fabrication. The system addresses the data inconsistency problem by one-hot encoding and data normalization techniques, Pearson correlation coefficient for feature selection, and the Random Forest Regression (RFR) for prediction. Compared with the other three Machine-Learning (ML) algorithms, the Random Forest algorithm has the best performance. The results demonstrate that the proposed system outperforms the conventional approach and has better forecast accuracy so it is suitable for man-hours prediction.

An effective segmentation and attention-based reptile residual capsule auto encoder for pest classification.

Insect pests are a major global factor affecting agricultural crop productivity and quality. Rapid and precise insect pest detection is crucial for improving handling and prediction techniques. There are several methods for pest detection and classification tasks; still, the inaccurate detection, computation complexity and several other challenges affect the performance of the model. Thus, this research presents a Deep Learning (DL) approach that has led to significant advancements and is currently being applied successfully in many domains, such as autonomous insect pest detection. Initially, the input images are gathered from the test dataset. The next step in pre-processing the input images is to improve the model capacity by removing unwanted data using the Enhanced Kuan filter method. Then, the pre-processed images are segmented using the Attention-based U-Net method. Finally, a novel Attention Based Reptile Residual Capsule Auto Encoder (ARRCAE) technique is proposed to classify and recognize crop pests. Furthermore, the Improved Reptile Search Optimisation (IRSO) algorithm is employed to fine-tune the classification parameters optimally. As a result, the proposed study enhances performance by classifying crop pest detection systems. The suggested method makes use of a Python tool for simulation, and pest datasets are utilized for result analysis. The suggested model beats other current models with an accuracy of 98%, precision of 97%, recall of 96%, and specificity of 99% for the pest dataset, per the simulation results that were obtained. © 2024 Society of Chemical Industry.

Floating-Point Embedding: Enhancing the Mathematical Comprehension of Large Language Models

The processing and comprehension of numerical information in natural language represent pivotal focal points of scholarly inquiry. Across diverse applications spanning text analysis to information retrieval, the adept management and understanding of the numerical content within natural language are indispensable in achieving task success. Specialized encoding and embedding techniques tailored to numerical data offer an avenue toward improved performance in tasks involving masked prediction and numerical reasoning, inherently characterized by numerical values. Consequently, treating numbers in text merely as words is inadequate; their numerical semantics must be underscored. Recent years have witnessed the emergence of a range of specific encoding methodologies designed explicitly for numerical content, demonstrating promising outcomes. We observe similarities between the Transformer architecture and CPU architecture, with symmetry playing a crucial role. In light of this observation and drawing inspiration from computer system theory, we introduce a floating-point representation and devise a corresponding embedding module. The numerical representations correspond one-to-one with their semantic vector values, rendering both symmetric regarding intermediate transformation methods. Our proposed methodology facilitates the more comprehensive encoding and embedding of numerical information within a predefined precision range, thereby ensuring a distinctive encoding representation for each numerical entity. Rigorous testing on multiple encoder-only models and datasets yielded results that stand out in terms of competitiveness. In comparison to the default embedding methods employed by models, our approach achieved an improvement of approximately 3.8% in Top-1 accuracy and a reduction in perplexity of approximately 0.43. These outcomes affirm the efficacy of our proposed method. Furthermore, the enrichment of numerical semantics through a more comprehensive embedding contributes to the augmentation of the model’s capacity for semantic understanding.

Switching Self-Attention Text Classification Model with Innovative Reverse Positional Encoding for Right-to-Left Languages: A Focus on Arabic Dialects

Transformer models have emerged as frontrunners in the field of natural language processing, primarily due to their adept use of self-attention mechanisms to grasp the semantic linkages between words in sequences. Despite their strengths, these models often face challenges in single-task learning scenarios, particularly when it comes to delivering top-notch performance and crafting strong latent feature representations. This challenge is more pronounced in the context of smaller datasets and is particularly acute for under-resourced languages such as Arabic. In light of these challenges, this study introduces a novel methodology for text classification of Arabic texts. This method harnesses the newly developed Reverse Positional Encoding (RPE) technique. It adopts an inductive-transfer learning (ITL) framework combined with a switching self-attention shared encoder, thereby increasing the model’s adaptability and improving its sentence representation accuracy. The integration of Mixture of Experts (MoE) and RPE techniques empowers the model to process longer sequences more effectively. This enhancement is notably beneficial for Arabic text classification, adeptly supporting both the intricate five-point and the simpler ternary classification tasks. The empirical evidence points to its outstanding performance, achieving accuracy rates of 87.20% for the HARD dataset, 72.17% for the BRAD dataset, and 86.89% for the LABR dataset, as evidenced by the assessments conducted on these datasets.

Speech-induced suppression during natural dialogues

When engaged in a conversation, one receives auditory information from the other’s speech but also from their own speech. However, this information is processed differently by an effect called Speech-Induced Suppression. Here, we studied brain representation of acoustic properties of speech in natural unscripted dialogues, using electroencephalography (EEG) and high-quality speech recordings from both participants. Using encoding techniques, we were able to reproduce a broad range of previous findings on listening to another’s speech, and achieving even better performances when predicting EEG signal in this complex scenario. Furthermore, we found no response when listening to oneself, using different acoustic features (spectrogram, envelope, etc.) and frequency bands, evidencing a strong effect of SIS. The present work shows that this mechanism is present, and even stronger, during natural dialogues. Moreover, the methodology presented here opens the possibility of a deeper understanding of the related mechanisms in a wider range of contexts.