Multi-class Classification Task Research Articles

Transforming two-dimensional tensor networks into quantum circuits for supervised learning

There have been numerous quantum neural networks reported, but they struggle to match traditional neural networks in accuracy. Given the huge improvement of the neural network models’ accuracy by two-dimensional tensor network (TN) states in classical tensor network machine learning (TNML), it is promising to explore whether its application in quantum machine learning can extend the performance boundary of the models. Here, we transform two-dimensional TNs into quantum circuits for supervised learning. Specifically, we encode two-dimensional TNs into quantum circuits through rigorous mathematical proofs for constructing model ansätze, including string-bond states, entangled-plaquette states and isometric TN states. In addition, we propose adaptive data encoding methods and combine with TNs. We construct a tensor-network-inspired quantum circuit (TNQC) supervised learning framework for transferring TNML from classical to quantum, and build several novel two-dimensional TN-inspired quantum classifiers based on this framework. Finally, we propose a parallel quantum machine learning method for multi-class classification to construct 2D TNQC-based multi-class classifiers. Classical simulation results on the MNIST benchmark dataset show that our proposed models achieve the state-of-the-art accuracy performance, significantly outperforming other quantum classifiers on both binary and multi-class classification tasks, and beat simple convolutional classifiers on a fair track with identical inputs. The noise resilience of the models makes them successfully run and work in a real quantum computer.

Unleashing the potential of fNIRS with machine learning: classification of fine anatomical movements to empower future brain-computer interface.

In this study, we explore the potential of using functional near-infrared spectroscopy (fNIRS) signals in conjunction with modern machine-learning techniques to classify specific anatomical movements to increase the number of control commands for a possible fNIRS-based brain-computer interface (BCI) applications. The study focuses on novel individual finger-tapping, a well-known task in fNIRS and fMRI studies, but limited to left/right or few fingers. Twenty-four right-handed participants performed the individual finger-tapping task. Data were recorded by using sixteen sources and detectors placed over the motor cortex according to the 10-10 international system. The event's average oxygenated Δ HbO and deoxygenated Δ HbR hemoglobin data were utilized as features to assess the performance of diverse machine learning (ML) models in a challenging multi-class classification setting. These methods include LDA, QDA, MNLR, XGBoost, and RF. A new DL-based model named "Hemo-Net" has been proposed which consists of multiple parallel convolution layers with different filters to extract the features. This paper aims to explore the efficacy of using fNRIS along with ML/DL methods in a multi-class classification task. Complex models like RF, XGBoost, and Hemo-Net produce relatively higher test set accuracy when compared to LDA, MNLR, and QDA. Hemo-Net has depicted a superior performance achieving the highest test set accuracy of 76%, however, in this work, we do not aim at improving the accuracies of models rather we are interested in exploring if fNIRS has the neural signatures to help modern ML/DL methods in multi-class classification which can lead to applications like brain-computer interfaces. Multi-class classification of fine anatomical movements, such as individual finger movements, is difficult to classify with fNIRS data. Traditional ML models like MNLR and LDA show inferior performance compared to the ensemble-based methods of RF and XGBoost. DL-based method Hemo-Net outperforms all methods evaluated in this study and demonstrates a promising future for fNIRS-based BCI applications.

Custom Loss Functions in XGBoost Algorithm for Enhanced Critical Error Mitigation in Drill-Wear Analysis of Melamine-Faced Chipboard.

The advancement of machine learning in industrial applications has necessitated the development of tailored solutions to address specific challenges, particularly in multi-class classification tasks. This study delves into the customization of loss functions within the eXtreme Gradient Boosting (XGBoost) algorithm, which is a critical step in enhancing the algorithm's performance for specific applications. Our research is motivated by the need for precision and efficiency in the industrial domain, where the implications of misclassification can be substantial. We focus on the drill-wear analysis of melamine-faced chipboard, a common material in furniture production, to demonstrate the impact of custom loss functions. The paper explores several variants of Weighted Softmax Loss Functions, including Edge Penalty and Adaptive Weighted Softmax Loss, to address the challenges of class imbalance and the heightened importance of accurately classifying edge classes. Our findings reveal that these custom loss functions significantly reduce critical errors in classification without compromising the overall accuracy of the model. This research not only contributes to the field of industrial machine learning by providing a nuanced approach to loss function customization but also underscores the importance of context-specific adaptations in machine learning algorithms. The results showcase the potential of tailored loss functions in balancing precision and efficiency, ensuring reliable and effective machine learning solutions in industrial settings.

Neural networks with functional inputs for multi-class supervised classification of replicated point patterns

AbstractA spatial point pattern is a collection of points observed in a bounded region of the Euclidean plane or space. With the dynamic development of modern imaging methods, large datasets of point patterns are available representing for example sub-cellular location patterns for human proteins or large forest populations. The main goal of this paper is to show the possibility of solving the supervised multi-class classification task for this particular type of complex data via functional neural networks. To predict the class membership for a newly observed point pattern, we compute an empirical estimate of a selected functional characteristic. Then, we consider such estimated function to be a functional variable entering the network. In a simulation study, we show that the neural network approach outperforms the kernel regression classifier that we consider a benchmark method in the point pattern setting. We also analyse a real dataset of point patterns of intramembranous particles and illustrate the practical applicability of the proposed method.

WITHDRAWN: Extracting key topics from massive COVID-19 information on social networks: An integrated deep learning and LDA framework

Quickly discovering and processing unmeaningful information is especially important in the current social network, where public opinion is rapidly fermenting. In this study, a novel topic extraction framework is proposed to extract key topics for the special event of COVID-19 on social networks. First, huge and noisy COVID-19 information is pre-processed for our framework. Second, a model of key information identification is designed for COVID-19 information, which combines the deep learning model of Text Convolutional Neural Networks (TextCNN) with Bi-directional Long Short-Term Memory (BiLSTM) to perform a multi-class classification task. Finally, a model of key topic extraction is developed with the Latent Dirichlet Allocation (LDA) model, which is able to summarize topics and keywords from a large amount of information. we crawled the public opinion information of the COVID-19 on weibo as our dataset. experimental results demonstrate that our model outperforms the other benchmark models in identifying important information related to COVID-19. the key topic extraction model based on LDA can further divide key information into multiple topics ►And describe the topics through keywords for fast tracking of COVID-19 pandemic on social networks.

Discriminative Regression With Adaptive Graph Diffusion.

In this article, we propose a new linear regression (LR)-based multiclass classification method, called discriminative regression with adaptive graph diffusion (DRAGD). Different from existing graph embedding-based LR methods, DRAGD introduces a new graph learning and embedding term, which explores the high-order structure information between four tuples, rather than conventional sample pairs to learn an intrinsic graph. Moreover, DRAGD provides a new way to simultaneously capture the local geometric structure and representation structure of data in one term. To enhance the discriminability of the transformation matrix, a retargeted learning approach is introduced. As a result of combining the above-mentioned techniques, DRAGD can flexibly explore more unsupervised information underlying the data and the label information to obtain the most discriminative transformation matrix for multiclass classification tasks. Experimental results on six well-known real-world databases and a synthetic database demonstrate that DRAGD is superior to the state-of-the-art LR methods.

Machine-Learning-Based Approaches for Multi-Level Sentiment Analysis of Romanian Reviews

Sentiment analysis has increasingly gained significance in commercial settings, driven by the rising impact of reviews on purchase decision-making in recent years. This research conducts a thorough examination of the suitability of machine learning and deep learning approaches for sentiment analysis, using Romanian reviews as a case study, with the aim of gaining insights into their practical utility. A comprehensive, multi-level analysis is performed, covering the document, sentence, and aspect levels. The main contributions of the paper refer to the in-depth exploration of multiple sentiment analysis models at three different textual levels and the subsequent improvements brought with respect to these standard models. Additionally, a balanced dataset of Romanian reviews from twelve product categories is introduced. The results indicate that, at the document level, supervised deep learning techniques yield the best outcomes (specifically, a convolutional neural network model that obtains an AUC value of 0.93 for binary classification and a weighted average F1-score of 0.77 in a multi-class setting with 5 target classes), albeit with increased resource consumption. Favorable results are achieved at the sentence level, as well, despite the heightened complexity of sentiment identification. In this case, the best-performing model is logistic regression, for which a weighted average F1-score of 0.77 is obtained in a multi-class polarity classification task with three classes. Finally, at the aspect level, promising outcomes are observed in both aspect term extraction and aspect category detection tasks, in the form of coherent and easily interpretable word clusters, encouraging further exploration in the context of aspect-based sentiment analysis for the Romanian language.

AVID: A speech database for machine learning studies on vocal intensity

Vocal intensity, which is quantified typically with the sound pressure level (SPL), is a key feature of speech. To measure SPL from speech recordings, a standard calibration tone (with a reference SPL of 94 dB or 114 dB) needs to be recorded together with speech. However, most of the popular databases that are used in areas such as speech and speaker recognition have been recorded without calibration information by expressing speech on arbitrary amplitude scales. Therefore, information about vocal intensity of the recorded speech, including SPL, is lost. In the current study, we introduce a new open and calibrated speech/electroglottography (EGG) database named Aalto Vocal Intensity Database (AVID). AVID includes speech and EGG produced by 50 speakers (25 males, 25 females) who varied their vocal intensity in four categories (soft, normal, loud and very loud). Recordings were conducted using a constant mouth-to-microphone distance and by recording a calibration tone. The speech data was labelled sentence-wise using a total of 19 labels that support the utilisation of the data in machine learning (ML) -based studies of vocal intensity based on supervised learning. In order to demonstrate how the AVID data can be used to study vocal intensity, we investigated one multi-class classification task (classification of speech into soft, normal, loud and very loud intensity classes) and one regression task (prediction of SPL of speech). In both tasks, we deliberately warped the level of the input speech by normalising the signal to have its maximum amplitude equal to 1.0, that is, we simulated a scenario that is prevalent in current speech databases. The results show that using the spectrogram feature with the support vector machine classifier gave an accuracy of 82% in the multi-class classification of the vocal intensity category. In the prediction of SPL, using the spectrogram feature with the support vector regressor gave an mean absolute error of about 2 dB and a coefficient of determination of 92%. We welcome researchers interested in classification and regression problems to utilise AVID in the study of vocal intensity, and we hope that the current results could serve as baselines for future ML studies on the topic.

Pose-based tremor type and level analysis for Parkinson's disease from video.

Current methods for diagnosis of PD rely on clinical examination. The accuracy of diagnosis ranges between 73 and 84%, and is influenced by the experience of the clinical assessor. Hence, an automatic, effective and interpretable supporting system for PD symptom identification would support clinicians in making more robust PD diagnostic decisions. We propose to analyze Parkinson's tremor (PT) to support the analysis of PD, since PT is one of the most typical symptoms of PD with broad generalizability. To realize the idea, we present SPA-PTA, a deep learning-based PT classification and severity estimation system that takes consumer-grade videos of front-facing humans as input. The core of the system is a novel attention module with a lightweight pyramidal channel-squeezing-fusion architecture that effectively extracts relevant PT information and filters noise. It enhances modeling performance while improving system interpretability. We validate our system via individual-based leave-one-out cross-validation on two tasks: the PT classification task and the tremor severity rating estimation task. Our system presents a 91.3% accuracy and 80.0% F1-score in classifying PT with non-PT class, while providing a 76.4% accuracy and 76.7% F1-score in more complex multiclass tremor rating classification task. Our system offers a cost-effective PT classification and tremor severity estimation results as warning signs of PD for undiagnosed patients with PT symptoms. In addition, it provides a potential solution for supporting PD diagnosis in regions with limited clinical resources.

DiCleave: a deep learning model for predicting human Dicer cleavage sites

BackgroundMicroRNAs (miRNAs) are a class of non-coding RNAs that play a pivotal role as gene expression regulators. These miRNAs are typically approximately 20 to 25 nucleotides long. The maturation of miRNAs requires Dicer cleavage at specific sites within the precursor miRNAs (pre-miRNAs). Recent advances in machine learning-based approaches for cleavage site prediction, such as PHDcleav and LBSizeCleav, have been reported. ReCGBM, a gradient boosting-based model, demonstrates superior performance compared with existing methods. Nonetheless, ReCGBM operates solely as a binary classifier despite the presence of two cleavage sites in a typical pre-miRNA. Previous approaches have focused on utilizing only a fraction of the structural information in pre-miRNAs, often overlooking comprehensive secondary structure information. There is a compelling need for the development of a novel model to address these limitations.ResultsIn this study, we developed a deep learning model for predicting the presence of a Dicer cleavage site within a pre-miRNA segment. This model was enhanced by an autoencoder that learned the secondary structure embeddings of pre-miRNA. Benchmarking experiments demonstrated that the performance of our model was comparable to that of ReCGBM in the binary classification tasks. In addition, our model excelled in multi-class classification tasks, making it a more versatile and practical solution than ReCGBM.ConclusionsOur proposed model exhibited superior performance compared with the current state-of-the-art model, underscoring the effectiveness of a deep learning approach in predicting Dicer cleavage sites. Furthermore, our model could be trained using only sequence and secondary structure information. Its capacity to accommodate multi-class classification tasks has enhanced the practical utility of our model.

Evidence Mining for Interpretable Charge Prediction via Prompt Learning

Nowadays, more and more researchers are committed to applying artificial intelligence technology to the legal field to support decision-making. Charge prediction is a subtask of legal judgment prediction (LJP). Its purpose is to analyze the fact description of natural language text and predict a charge corresponding to a case. At present, most of the research takes charge prediction as a multiclass classification task, which leads to unsatisfactory interpretation due to the weak semantic correlation between the fact descriptions and charge labels. In order to solve this problem, we propose a method of generative evidence mining based on prompt learning. Specifically, in the training phase, we reformulate the charge labels into the prompt template that we design to enhance the semantic correlation between the charge labels and the fact descriptions. In the testing phase, the charge labels are generated via the model based on prompt learning. Meanwhile, we calculate the attention score of each sentence from the multihead self-attention in the transformer encoder and choose the sentence with the highest attention score as the evidence. Our experimental results on a real dataset show that our method is better than the traditional fine-tuning-based classification method.

VGG19 Demonstrates the Highest Accuracy Rate in a Nine-Class Wound Classification Task Among Various Deep Learning Networks: A Pilot Study

Background. Current literature suggests relatively low accuracy of multi-class wound classification tasks using deep learning networks. Solutions are needed to address the increasing diagnostic burden of wounds on wound care professionals and to aid non-wound care professionals in wound management. Objective. To develop a reliable, accurate 9-class classification system to aid wound care professionals and perhaps eventually, patients and non-wound care professionals, in managing wounds. Methods. A total of 8173 training data images and 904 test data images were classified into 9 categories: operation wound, laceration, abrasion, skin defect, infected wound, necrosis, diabetic foot ulcer, chronic ulcer, and wound dehiscence. Six deep learning networks, based on VGG16, VGG19, EfficientNet-B0, EfficientNet-B5, RepVGG-A0, and RepVGG-B0, were established, trained, and tested on the same images. For each network the accuracy rate, defined as the sum of true positive and true negative values divided by the total number, was analyzed. Results. The overall accuracy varied from 74.0% to 82.4%. Of all the networks, VGG19 achieved the highest accuracy, at 82.4%. This result is comparable to those reported in previous studies. Conclusion. These findings indicate the potential for VGG19 to be the basis for a more comprehensive and detailed AI-based wound diagnostic system. Eventually, such systems also may aid patients and non-wound care professionals in diagnosing and treating wounds.

Task-Related Saliency for Few-Shot Image Classification.

A weakness of the existing metric-based few-shot classification method is that task-unrelated objects or backgrounds may mislead the model since the small number of samples in the support set is insufficient to reveal the task-related targets. An essential cue of human wisdom in the few-shot classification task is that they can recognize the task-related targets by a glimpse of support images without being distracted by task-unrelated things. Thus, we propose to explicitly learn task-related saliency features and make use of them in the metric-based few-shot learning schema. We divide the tackling of the task into three phases, namely, the modeling, the analyzing, and the matching. In the modeling phase, we introduce a saliency sensitive module (SSM), which is an inexact supervision task jointly trained with a standard multiclass classification task. SSM not only enhances the fine-grained representation of feature embedding but also can locate the task-related saliency features. Meanwhile, we propose a self-training-based task-related saliency network (TRSN) which is a lightweight network to distill task-related salience produced by SSM. In the analyzing phase, we freeze TRSN and use it to handle novel tasks. TRSN extracts task-relevant features while suppressing the disturbing task-unrelated features. We, therefore, can discriminate samples accurately in the matching phase by strengthening the task-related features. We conduct extensive experiments on five-way 1-shot and 5-shot settings to evaluate the proposed method. Results show that our method achieves a consistent performance gain on benchmarks and achieves the state-of-the-art.

Supervised Contrastive Learning Framework and Hardware Implementation of Learned ResNet for Real-time Respiratory Sound Classification.

This paper presents a supervised contrastive learning (SCL) framework for respiratory sound classification and the hardware implementation of learned ResNet on field programmable gate array (FPGA) for real-time monitoring. At the algorithmic level, multiple techniques such as features augmentation and MixUp are combined holistically to mitigate the impact of data scarcity and imbalanced classes in the training dataset. Bayesian optimization further enhances the classification accuracy through parameter tuning in pre-processing and SCL. The proposed framework achieves 0.8725 total score (including runtime score) on a ResNet-18 model in both event and record multi-class classification tasks using the SJTU Paediatric Respiratory Sound Database (SPRSound). In addition, algorithm-hardware co-optimizations including Quantization-Aware Training (QAT), merge of network layers, optimization of memory size and number of parallel threads are performed for hardware implementation on FPGA. This approach reduces 40% model size and 70% computation latency. The learned ResNet is implemented on a Xilinx Zynq ZCU102 FPGA with 16ms latency and less than 2% inference score degradation compared to the software model.

Spectrum-based channel attention cooperating with time continuity encoding in transformer for EEG emotion analysis

Recently, electroencephalogram (EEG) emotion analysis has attracted increasing attention in many fields, such as neuroscience and brain-computer fusion. Due to the spatial channel difference and time continuity of emotional responses, it is necessary to capture relevant properties from EEG signals. In this paper, spectrum-based spatial channel attention is developed to reveal the emotional responses of different electrode channels located in different cerebral cortex regions. In addition, a time continuity encoding mechanism is developed to encode the temporal relations of time series signals in the transformer. In our model, the spectrum-based spatial channel attention cooperates with the time continuity encoding mechanism to fully utilize the spectral, spatial and temporal features of EEG signals (SST-Emo). Extensive experiments with different window sizes (1 s and 3 s) are carried out on DEAP for both binary and multiclass classification tasks. The accuracies are 96.28 %, 95.25 %, 95.52 %, 95.83 % and 93.88 % for 1 s window-size on arousal, valence, dominance, liking and 5-class, respectively. For 3 s window-size, the accuracies are 96.13 %, 95.74 %, 95.23 %, 95.7 % and 95.83 %, respectively. The results show that SST-Emo is superior to the state-of-the-art methods, which indicates that SST-Emo is able to extract more discriminative emotional representations.

Artificial Intelligence on Diagnostic Aid of Leprosy: A Systematic Literature Review.

Leprosy is a neglected tropical disease that can cause physical injury and mental disability. Diagnosis is primarily clinical, but can be inconclusive due to the absence of initial symptoms and similarity to other dermatological diseases. Artificial intelligence (AI) techniques have been used in dermatology, assisting clinical procedures and diagnostics. In particular, AI-supported solutions have been proposed in the literature to aid in the diagnosis of leprosy, and this Systematic Literature Review (SLR) aims to characterize the state of the art. This SLR followed the preferred reporting items for systematic reviews and meta-analyses (PRISMA) framework and was conducted in the following databases: ACM Digital Library, IEEE Digital Library, ISI Web of Science, Scopus, and PubMed. Potentially relevant research articles were retrieved. The researchers applied criteria to select the studies, assess their quality, and perform the data extraction process. Moreover, 1659 studies were retrieved, of which 21 were included in the review after selection. Most of the studies used images of skin lesions, classical machine learning algorithms, and multi-class classification tasks to develop models to diagnose dermatological diseases. Most of the reviewed articles did not target leprosy as the study's primary objective but rather the classification of different skin diseases (among them, leprosy). Although AI-supported leprosy diagnosis is constantly evolving, research in this area is still in its early stage, then studies are required to make AI solutions mature enough to be transformed into clinical practice. Expanding research efforts on leprosy diagnosis, coupled with the advocacy of open science in leveraging AI for diagnostic support, can yield robust and influential outcomes.

Morphology-based deep learning approach for predicting adipogenic and osteogenic differentiation of human mesenchymal stem cells (hMSCs)

Human mesenchymal stem cells (hMSCs) are multipotent progenitor cells with the potential to differentiate into various cell types, including osteoblasts, chondrocytes, and adipocytes. These cells have been extensively employed in the field of cell-based therapies and regenerative medicine due to their inherent attributes of self-renewal and multipotency. Traditional approaches for assessing hMSCs differentiation capacity have relied heavily on labor-intensive techniques, such as RT-PCR, immunostaining, and Western blot, to identify specific biomarkers. However, these methods are not only time-consuming and economically demanding, but also require the fixation of cells, resulting in the loss of temporal data. Consequently, there is an emerging need for a more efficient and precise approach to predict hMSCs differentiation in live cells, particularly for osteogenic and adipogenic differentiation. In response to this need, we developed innovative approaches that combine live-cell imaging with cutting-edge deep learning techniques, specifically employing a convolutional neural network (CNN) to meticulously classify osteogenic and adipogenic differentiation. Specifically, four notable pre-trained CNN models, VGG 19, Inception V3, ResNet 18, and ResNet 50, were developed and tested for identifying adipogenic and osteogenic differentiated cells based on cell morphology changes. We rigorously evaluated the performance of these four models concerning binary and multi-class classification of differentiated cells at various time intervals, focusing on pivotal metrics such as accuracy, the area under the receiver operating characteristic curve (AUC), sensitivity, precision, and F1-score. Among these four different models, ResNet 50 has proven to be the most effective choice with the highest accuracy (0.9572 for binary, 0.9474 for multi-class) and AUC (0.9958 for binary, 0.9836 for multi-class) in both multi-class and binary classification tasks. Although VGG 19 matched the accuracy of ResNet 50 in both tasks, ResNet 50 consistently outperformed it in terms of AUC, underscoring its superior effectiveness in identifying differentiated cells. Overall, our study demonstrated the capability to use a CNN approach to predict stem cell fate based on morphology changes, which will potentially provide insights for the application of cell-based therapy and advance our understanding of regenerative medicine.

Generating Synthetic Resume Data with Large Language Models for Enhanced Job Description Classification

In this article, we investigate the potential of synthetic resumes as a means for the rapid generation of training data and their effectiveness in data augmentation, especially in categories marked by sparse samples. The widespread implementation of machine learning algorithms in natural language processing (NLP) has notably streamlined the resume classification process, delivering time and cost efficiencies for hiring organizations. However, the performance of these algorithms depends on the abundance of training data. While selecting the right model architecture is essential, it is also crucial to ensure the availability of a robust, well-curated dataset. For many categories in the job market, data sparsity remains a challenge. To deal with this challenge, we employed the OpenAI API to generate both structured and unstructured resumes tailored to specific criteria. These synthetically generated resumes were cleaned, preprocessed and then utilized to train two distinct models: a transformer model (BERT) and a feedforward neural network (FFNN) that incorporated Universal Sentence Encoder 4 (USE4) embeddings. While both models were evaluated on the multiclass classification task of resumes, when trained on an augmented dataset containing 60 percent real data (from Indeed website) and 40 percent synthetic data from ChatGPT, the transformer model presented exceptional accuracy. The FFNN, albeit predictably, achieved lower accuracy. These findings highlight the value of augmented real-world data with ChatGPT-generated synthetic resumes, especially in the context of limited training data. The suitability of the BERT model for such classification tasks further reinforces this narrative.

Advancement in Bangla Sentiment Analysis: A Comparative Study of Transformer-Based and Transfer Learning Models for E-commerce Sentiment Classification

Background: As a direct result of the Internet's expansion, the quantity of information shared by Internet users across its numerous platforms has increased. Sentiment analysis functions at a higher level when there are more available perspectives and opinions. However, the lack of labeled data significantly complicates sentiment analysis utilizing Bangla natural language processing (NLP). In recent years, nevertheless, due to the development of more effective deep learning models, Bangla sentiment analysis has improved significantly. Objective: This article presents a curated dataset for Bangla e-commerce sentiment analysis obtained solely from the "Daraz" platform. We aim to conduct sentiment analysis in Bangla for binary and understudied multiclass classification tasks. Methods: Transfer learning (LSTM, GRU) and Transformers (Bangla-BERT) approaches are compared for their effectiveness on our dataset. To enhance the overall performance of the models, we fine-tuned them. Results: The accuracy of Bangla-BERT was highest for both binary and multiclass sentiment classification tasks, with 94.5% accuracy for binary classification and 88.78% accuracy for multiclass sentiment classification. Conclusion: Our proposed method performs noticeably better classifying multiclass sentiments in Bangla than previous deep learning techniques. Keywords: Bangla-BERT, Deep Learning, E-commerce, NLP, Sentiment Analysis

Accuracy calibration transfer between multi-sensor systems based on metal-oxide gas sensors used for analysis of exhaled air

The application of electronic nose system as an instrument for exhaled air analysis for diagnosing various diseases (lung cancer, etc.) is one of the most rapidly developing areas of predictive medicine. However, a significant disadvantage of such systems is difficulty in production of identical multi-sensor systems because the sensors of the same type are not identical, that degrades the predictive ability. To solve this problem calibration transfer methods are used.This paper analyzes currently applied calibration transfer methods. A new approach to evaluate the effectiveness of calibration transfer methods for two multi-sensor systems is proposed and implemented. One- and three-component gas mixtures are taken for the analysis. In this work, four methods are used to adjust the sensor responses: UDS, UDSwoi, DS-L1R and DS-PLS2. Applicability of one-factor standardization methods (UDS and UDSwoi) is shown on the classification tasks: multi-class classification task of one-component gas mixtures and binary classification task of three-component gas mixtures. Multi-class classification of model one-component gas mixtures for UDS, UDSwoi with 4 calibration transfer samples provided the classification accuracy of 0.985 and 0.990, respectively. The binary classification of two three-component mixtures with different ratios of three VOCs within the 20% discrepancy limits, UDS, UDSwoi and DS-PLS2 achieved an error-free accuracy of 1.000.