Macro F1 Score Research Articles

Enhancing Network Threat Detection with Random Forest-Based NIDS and Permutation Feature Importance

Network Intrusion Detection Systems (NIDS) are critical for protecting computer networks from unauthorized activities. Traditional NIDS rely on rule-based signatures, which can be limiting in detecting emerging threats. This study investigates the effectiveness of the random forest classifier in advancing NIDS capabilities through machine learning. Using the CICIDS-2017 dataset, the data are preprocessed to enhance their quality by removing redundancies. feature selection and permutation importance were employed to identify the most relevant features. The methodology involves rigorous testing and analysis of the random forest classifier’s performance, focusing on f1-score rates compared to other machine learning models. Results demonstrate that by optimizing class weights, applying a custom prediction function and leveraging 26 key features, the random forest classifier achieves an outstanding 99.8% in the weighted f1-score and 93.31% in the macro f1-score in various attack types. This research highlights the potential of machine learning to significantly enhance NIDS effectiveness, offering a robust defense mechanism against evolving cybersecurity threats in modern networks.

Classification of Corn Diseases and Pests Using Fuzzy Naïve Bayes Method

Corn is an essential dietary source for humans and animals. In addition to being a food source, corn has numerous benefits as a manufacturing commodity. The quality of grain crops must be considered to minimise the likelihood of disease and pest infestations. Therefore, the diseases and pests that attack corn plants must be classified so that farmers can control them during the growth period of corn plants. The fuzzy naive Bayes method is a statistical machine learning method that can be used to classify the diseases and pests of corn crops based on colour space-transformed digital images. This study aims to classify corn plant diseases and pests using the fuzzy naive Bayes method. Digital images of corn plant diseases and pests were transformed into a red, green and blue colour space model. The following seven classes of corn plant diseases and pests were classified: leaf rust disease, downy mildew disease, leaf blight disease, Locusta pest, Heliotis armigera pest, Spodoptera frugiperdita pest and non-pathogenic pest. With this method, the classification model achieves an accuracy of 87.83%, a macro precision of 34.91%, a macro recall of 35.90% and a macro f-score of 33.82%.

Identify devices and events from non-IP heterogeneous IoT network traffic

In recent years, notable advancements have been achieved in the realm of identifying IP-based Internet of Things (IoT) devices and events. Nevertheless, the majority of methods rely on extracting fingerprints or features from plain text IP-based packets, which limits their ability to accommodate heterogeneous IoT devices such as ZigBee and Z-Wave, and fails to address the challenge of limited traffic samples. To tackle these issues, we propose a novel approach based on IoT communication characteristics and featuring module extensibility. This method is presented to effectively identify IoT devices and events from non-IP heterogeneous IoT network traffic. To shield the differences caused by the heterogeneous IoT protocol, a heterogeneous sample extraction platform with an extensible structure is created to extract raw sequence samples from ZigBee and Z-Wave traffic, with potential for expansion to other protocols. To address the challenges arising from the scarcity of samples, a sample identification framework based on IoT communication characteristics is devised to create synthetic samples from the raw sequence samples, enabling concurrent processing of the raw and synthetic samples using an identification model featuring two separate sequence networks. Comparative assessments of our method against baseline sequence models and the latest techniques demonstrate the advantages of our approach in identifying non-IP heterogeneous IoT traffic. The experimental results indicate that our method achieves an average accuracy improvement of 29.7% compared to baseline models using only raw samples. Furthermore, our method shows improvements of 22.1%, 21.5%, and 21.8% in macro precision, macro recall, and macro F1-score, respectively, over the latest method.

GASCOM: Graph-based Attentive Semantic Context Modeling for Online Conversation Understanding

Online conversation understanding is an important yet challenging NLP problem which has many useful applications (e.g., hate speech detection). However, online conversations typically unfold over a series of posts and replies to those posts, forming a tree structure within which individual posts may refer to semantic context from elsewhere in the tree. Such semantic cross-referencing makes it difficult to understand a single post by itself; yet considering the entire conversation tree is not only difficult to scale but can also be misleading as a single conversation may have several distinct threads or points, not all of which are relevant to the post being considered. In this paper, we propose a Graph-based Attentive Semantic COntext Modeling (GASCOM) framework for online conversation understanding. Specifically, we design two novel algorithms that utilize both the graph structure of the online conversation as well as the semantic information from individual posts for retrieving relevant context nodes from the whole conversation. We further design a token-level multi-head graph attention mechanism to pay different attentions to different tokens from different selected context utterances for fine-grained conversation context modelling. Using this semantic conversational context, we re-examine two well-studied problems: polarity prediction and hate speech detection. Our proposed framework significantly outperforms state-of-the-art methods on both tasks, improving macro-F1 scores by 4.5% for polarity prediction and by 5% for hate speech detection. The GASCOM context weights also enhance interpretability.

A Comparative Study of Transformer-based Models for Hate-Speech Detection in English-Kiswahili Code-Switched Social Media Text

The transformer architecture, first introduced in 2017 by researchers at Google, has revolutionized natural language processing in various tasks, including text classification. This architecture formed the basis of future models such as those used in hate speech detection in code-switched text. In this research, we conduct a comparative study of transformer-based models for hate speech detection in English-Kiswahili code-switched text. First, the models were compared as feature extractors using a traditional classifier and then as end-to-end classifiers. The three multilingual transformer-based models compared include mBERT, mDistilBERT and XLM-RoBERTa, using SVM as the traditional classifier for the extracted features. The HateSpeech_Kenya dataset, sourced from Kaggle, was utilized in this study. As a feature extractor, mBERT’s hidden states trained the highest-performing SVM with an accuracy of 0.5461 and a macro f1 score of 0.40. Among the three models evaluated, XLM-RoBERTa achieved the highest accuracy of 0.6069 and a macro f1 score of 0.49 on a balanced dataset. In contrast, mBERT achieved the highest accuracy of 0.7820 and a macro f1 score of 0.53 on an imbalanced dataset. The comparative study establishes that using transformer-based models as end-to-end classifiers generally performs better than using them as feature extractors with traditional classifiers. This is because directly training the models allows them to learn more task-specific features. Furthermore, the varying performance across balanced and imbalanced datasets highlights the need for careful model selection based on the dataset characteristics and specific task requirements.

Large-scale identification of social and behavioral determinants of health from clinical notes: comparison of Latent Semantic Indexing and Generative Pretrained Transformer (GPT) models

BackgroundSocial and behavioral determinants of health (SBDH) are associated with a variety of health and utilization outcomes, yet these factors are not routinely documented in the structured fields of electronic health records (EHR). The objective of this study was to evaluate different machine learning approaches for detection of SBDH from the unstructured clinical notes in the EHR.MethodsLatent Semantic Indexing (LSI) was applied to 2,083,180 clinical notes corresponding to 46,146 patients in the MIMIC-III dataset. Using LSI, patients were ranked based on conceptual relevance to a set of keywords (lexicons) pertaining to 15 different SBDH categories. For Generative Pretrained Transformer (GPT) models, API requests were made with a Python script to connect to the OpenAI services in Azure, using gpt-3.5-turbo-1106 and gpt-4-1106-preview models. Prediction of SBDH categories were performed using a logistic regression model that included age, gender, race and SBDH ICD-9 codes.ResultsLSI retrieved patients according to 15 SBDH domains, with an overall average PPV ≥\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\ge$$\\end{document} 83%. Using manually curated gold standard (GS) sets for nine SBDH categories, the macro-F1 score of LSI (0.74) was better than ICD-9 (0.71) and GPT-3.5 (0.54), but lower than GPT-4 (0.80). Due to document size limitations, only a subset of the GS cases could be processed by GPT-3.5 (55.8%) and GPT-4 (94.2%), compared to LSI (100%). Using common GS subsets for nine different SBDH categories, the macro-F1 of ICD-9 combined with either LSI (mean 0.88, 95% CI 0.82-0.93), GPT-3.5 (0.86, 0.82-0.91) or GPT-4 (0.88, 0.83-0.94) was not significantly different. After including age, gender, race and ICD-9 in a logistic regression model, the AUC for prediction of six out of the nine SBDH categories was higher for LSI compared to GPT-4.0.ConclusionsThese results demonstrate that the LSI approach performs comparable to more recent large language models, such as GPT-3.5 and GPT-4.0, when using the same set of documents. Importantly, LSI is robust, deterministic, and does not have document-size limitations or cost implications, which make it more amenable to real-world applications in health systems.

Revealing the impact of social circumstances on the selection of cancer therapy through natural language processing of social work notes.

We aimed to investigate the impact of social circumstances on cancer therapy selection using natural language processing to derive insights from social worker documentation. We developed and employed a Bidirectional Encoder Representations from Transformers (BERT) based approach, using a hierarchical multi-step BERT model (BERT-MS), to predict the prescription of targeted cancer therapy to patients based solely on documentation by clinical social workers. Our corpus included free-text clinical social work notes, combined with medication prescription information, for all patients treated for breast cancer at UCSF between 2012 and 2021. We conducted a feature importance analysis to identify the specific social circumstances that impact cancer therapy regimen. Using only social work notes, we consistently predicted the administration of targeted therapies, suggesting systematic differences in treatment selection exist due to non-clinical factors. The findings were confirmed by several language models, with GatorTron achieving the best performance with an area under the receiver operating characteristic curve (AUROC) of 0.721 and a Macro F1 score of 0.616. The UCSF BERT-MS model, capable of leveraging multiple pieces of notes, surpassed the UCSF-BERT model in both AUROC and Macro-F1. Our feature importance analysis identified several clinically intuitive social determinants of health that potentially contribute to disparities in treatment. Leveraging social work notes can be instrumental in identifying disparities in clinical decision-making. Hypotheses generated in an automated way could be used to guide patient-specific quality improvement interventions. Further validation with diverse clinical outcomes and prospective studies is essential. Our findings indicate that significant disparities exist among breast cancer patients receiving different types of therapies based on social determinants of health. Social work reports play a crucial role in understanding these disparities in clinical decision-making.

MeSHelper: Predicting the evolution of Medical Subject Headings based on knowledge graph dynamics

The Medical Subject Headings (MeSH) thesaurus is a controlled vocabulary widely used in biomedical knowledge systems. We propose a novel framework, termed MeSHelper, that employs a dynamic knowledge graph to predict whether the MeSH main headings (MHs) will evolve while also predicting their corresponding revision type. We parsed the whole PubMed database and all MeSH releases to construct a dynamic semantic tree (DST) and a dynamic knowledge network (DKN) to characterise the evolutionary patterns of MHs and create prediction models. Our results show that DST-related features play a major role in predicting whether the MHs will be revised. Our prediction performance achieved an F1 score of 92.07%. Both DST- and DKN-related features play a crucial role in predicting which types of MHs will evolve. The prediction performance achieved a Macro-F1 score of 72.15%, a Micro-F1 score of 84.09% and a Weighted-F1 score of 84.55%. The findings of this work aid both in constructing an automatic update model for domain thesauruses and in detecting evolutionary trends of the domain knowledge system.

Unravelling sleep patterns: Supervised contrastive learning with self-attention for sleep stage classification

Sleep data scoring is a crucial and primary step for diagnosing sleep disorders to know the sleep stages from the PSG signals. This study uses supervised contrastive learning with a self-attention mechanism to classify sleep stages. We propose a deep learning framework for automatic sleep stage classification, which involves two training phases: (1) the feature representation learning phase, in which the feature representation network (encoder) learns to extract features from the electroencephalogram (EEG) signals, and (2) the classification network training phase, where a pre-trained encoder (trained during phase I) along with the classifier head is fine-tuned for the classification task. The PSG data shows a non-uniform distribution of sleep stages, with wake (W) (around 30% of total samples) and N2 stages (around 58% and 37% of total samples in Physionet EDF-Sleep 2013 and 2018 datasets, respectively) being more prevalent, leading to an imbalanced dataset. The imbalanced data issue is addressed using a weighted softmax cross-entropy loss function that assigns higher weights to minority sleep stages. Additionally, an oversampling technique (the synthetic minority oversampling technique (SMOTE) (Chawla et al., 2002)[1] ) is applied to generate synthetic samples for minority classes. The proposed model is evaluated on the Physionet EDF-Sleep 2013 and 2018 datasets using Fpz-Cz and Pz-Oz EEG channels. It achieved an overall accuracy of 94.1%, a macro F1 score of 92.64, and a Cohen’s Kappa coefficient of 0.92. Ablation studies demonstrated the importance of triplet loss-based pre-training and oversampling for enhancing performance. The proposed model requires minimal pre-processing, eliminating the need for extensive signal processing expertise, and thus is well-suited for clinicians diagnosing sleep disorders.

Text mining approach for feature extraction and cartilage disease grade classification using knee MRI radiology reports

MRI radiology reporting processes can be improved by exploiting structured and semantically labelled data that can be fed to artificial intelligence (AI) tools. AI-based tools assisting radiology reporting can help to automatically individuate cartilage grading in textual magnetic resonance imaging (MRI) reports, thus supporting clinicians' decisions regarding medical imaging utilisation, diagnosis and treatment. In this study, we extracted information (clinical findings, observations, anatomical regions, etc.) and classified knee cartilage degradation from medical reports utilising transfer-learning techniques applied to the Bidirectional Encoder Representations from Transformers (BERT) model and its variants, pre-trained on an Italian-language corpus. To realise this objective, we used a dataset of 750 MRI knee reports written by three radiologists who contributed to a manual annotation process to perform text classification (TC) and named entity recognition (NER) tasks. The dataset was obtained from an internal database of the IRCCS SYNLAB SDN. Seventy percent of the dataset was used for training, 10% was used for validation and 20% was used for testing. The best-performing configurations for NER and TC tasks were based on the pre-trained BERT model. The macro F1-scores obtained with the NER and TC models are 0.89 and 0.81, respectively. The accuracies calculated on the test set for both tasks are 0.96 and 0.99, respectively.

CARROT: Simultaneous prediction of anomalies from groups of correlated cryptocurrency trends

Cryptocurrencies are virtual currencies that exploit cryptography to perform secure financial transactions. They gained widespread popularity in recent years due to their decentralized nature, (pseudo-)anonymity, and ability to facilitate cross-border transactions without the need for intermediaries. However, their price on the market exhibits a huge volatility, that makes them prone to market anomalies. Therefore, predicting anomalies in cryptocurrency time series can be considered an important task for financial institutions, traders, and investors, to maximize their profit or minimize losses.In this paper, we propose a novel approach for predicting anomalies in cryptocurrency time series by exploiting temporal correlations among different cryptocurrencies. Our approach, called CARROT, is based on the idea that groups of cryptocurrencies exhibit similar trends, possibly due to common influencing factors. CARROT analyzes the temporal correlation between different cryptocurrencies, and identifies clusters showing similar patterns that can be useful for gaining insights into future anomalies. Subsequently, CARROT exploits multiple (i.e., one for each cluster) multi-target LSTM models to predict anomalies.Our experiments, performed on a dataset of 17 cryptocurrencies, proved that CARROT outperforms single-target LSTM models of up to 20%, as well as other approaches based on neural networks, i.e., MLP and CNN, in terms of macro F1-score. Therefore, the proposed approach can be considered as a promising tool for predicting anomalies in cryptocurrency time series data and can potentially be used to improve risk management and trading strategies in the cryptocurrency market.

Automatic sleep staging based on 24/7 EEG SubQ (UNEEG medical) data displays strong agreement with polysomnography in healthy adults

Goal and aimsPerformance evaluation of automatic sleep staging on two-channel subcutaneous electroencephalography. Focus technologyUNEEG medical’s 24/7 electroencephalography SubQ (the SubQ device) with deep learning model U-SleepSQ. Reference method/technologyManually scored hypnograms from polysomnographic recordings. SampleTwenty-two healthy adults with 1-6 recordings per participant. The clinical study was registered at ClinicalTrials.gov with the identifier NCT04513743. DesignFine-tuning of U-Sleep in 11-fold cross-participant validation on 22 healthy adults. The resultant model was called U-SleepSQ. Core analyticsBland-Altman analysis of sleep parameters. Advanced multiclass model performance metrics: stage-specific accuracy, specificity, sensitivity, kappa (κ), and F1 score. Additionally, Cohen’s κ coefficient and macro F1 score. Longitudinal and participant-level performance evaluation. Additional analytics and exploratory analysesExploration of model confidence quantification. Performance vs. age, sex, body mass index, SubQ implantation hemisphere, normalized entropy, transition index, and scores from the following three questionnaires: Morningness-Eveningness Questionnaire, World Health Organization’s 5-item Well-being Index, and Major Depression Inventory. Core outcomesThere was a strong agreement between the focus and reference method/technology. Important supplemental outcomesThe confidence score was a promising metric for estimating the reliability of each hypnogram classified by the system. Core conclusionThe U-SleepSQ model classified hypnograms for healthy participants soon after implantation and longitudinally with a strong agreement with the gold standard of manually scored polysomnographics, exhibiting negligible temporal variation.

Text Analytics on YouTube Comments for Food Products

YouTube is a popular social media platform in the contemporary digital landscape. The primary focus of this study is to explore the underlying sentiment in user comments about food-related videos on YouTube, specifically within two pivotal food categories: plant-based and hedonic product. We labeled comments using sentiment lexicons such as TextBlob, VADER, and Google’s Sentiment Analysis (GSA) engine. Comment sentiment was classified using advanced Machine-Learning (ML) algorithms, namely Support Vector Machines (SVM), Multinomial Naive Bayes, Random Forest, Logistic Regression, and XGBoost. The evaluation of these models encompassed key macro average metrics, including accuracy, precision, recall, and F1 score. The results from GSA showed a high accuracy level, with SVM achieving 93% accuracy in the plant-based dataset and 96% in the hedonic dataset. In addition to sentiment analysis, we delved into user interactions within the two datasets, measuring crucial metrics, such as views, likes, comments, and engagement rate. The findings illuminate significantly higher levels of views, likes, and comments in the hedonic food dataset, but the plant-based dataset maintains a superior overall engagement rate.

Electro-Stimulation System with Artificial-Intelligence-Based Auricular-Triggered Algorithm to Support Facial Movements in Peripheral Facial Palsy: A Simulation Pilot Study.

Facial palsy causes severe functional disorders and impairs quality of life. Disturbing challenges for patients with acute facial palsy, but also with those with chronic facial palsy with synkinesis, are the loss of the ability to smile and insufficient eyelid closure. A potential treatment for these conditions could be a closed-loop electro-stimulation system that stimulates the facial muscles on the paretic side as needed to elicit eye closure, eye blink and smile in a manner similar to the healthy side. This study focuses on the development and evaluation of such a system. An artificial intelligence (AI)-based auricular-triggered algorithm is used to classify the intended facial movements. This classification is based on surface electromyography (EMG) recordings of the extrinsic auricular muscles, specifically the anterior, superior, and posterior auricular muscle on the paretic side. The system then delivers targeted surface electrical stimulation to contract the appropriate facial muscles. The evaluation of the system was conducted with 17 patients with facial synkinesis, who performed various facial movements according to a paradigm video. The system's performance was evaluated through a simulation, using previously captured data as the inputs. The performance was evaluated by means of the median macro F1-score, which was calculated based on the stimulation signal (output of the system) and the actual movements the patients performed. This study showed that such a system, using an AI-based auricular-triggered algorithm, can support with a median macro F1-score of 0.602 for the facial movements on the synkinetic side in patients with unilateral chronic facial palsy with synkinesis.

Multimodal reaching-position prediction for ADL support using neural networks

This study aimed to develop daily living support robots for patients with hemiplegia and the elderly. To support the daily living activities using robots in ordinary households without imposing physical and mental burdens on users, the system must detect the actions of the user and move appropriately according to their motions. We propose a reaching-position prediction scheme that targets the motion of lifting the upper arm, which is burdensome for patients with hemiplegia and the elderly in daily living activities. For this motion, it is difficult to obtain effective features to create a prediction model in environments where large-scale sensor system installation is not feasible and the motion time is short. We performed motion-collection experiments, revealed the features of the target motion and built a prediction model using the multimodal motion features and deep learning. The proposed model achieved an accuracy of 93% macro average and F1-score of 0.69 for a 9-class classification prediction at 35% of the motion completion.

A novel method for rice identification: Coupling Raman spectroscopy with Fourier spectrum and analyzing with deep learning

The development of detection methods for food adulteration is of great significance for promoting innovation and development in food safety supervision technology. In this study, we proposed a novel model: Raman-Fourier-BiLSTM-CNN (RFBC), which combines Raman spectroscopy with deep learning to achieve precise identification of seven brands of japonica rice in the northeastern region of China. This model focuses on revealing the worth of the Fourier spectrum of Raman spectra. The raw Raman spectra and their Fourier spectra are cleverly coupled through a Bi-directional Long Short-Term Memory (BiLSTM) connection structure, and the resulting composite features are deeply analyzed by a Convolutional Neural Network (CNN). Compared to traditional algorithms, RFBC exhibits superior feature extraction capabilities. To further evaluate the recognition performance of RFBC, this study compared it with the RFBC model containing only the original Raman spectrum branch, as well as with machine learning models based on Principal Component Analysis (PCA), including Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and eXtreme Gradient Boosting (XGBoost). The results show that, compared with the other five models, RFBC can accurately identify different brands of japonica rice, with significant advantages in accuracy, precision, recall, and other aspects. RFBC achieves a classification accuracy of 97.1 %, macro precision of 97.2 %, macro recall of 97.1 %, and macro F1-score of 97.2 %. The system proposed in this study can more accurately identify brands of japonica rice, providing strong technical support for combating counterfeit japonica rice products.

Efficient diagnostic classification of diverse pathologies through contextual eye movement data analysis with a novel hybrid architecture

The analysis of eye movements has proven valuable for understanding brain function and the neuropathology of various disorders. This research aims to utilize eye movement data analysis as a screening tool for differentiation between eight different groups of pathologies, including scholar, neurologic, and postural disorders. Leveraging a dataset from 20 clinical centers, all employing AIDEAL and REMOBI eye movement technologies this study extends prior research by considering a multi-annotation setting, incorporating information from recordings from saccade and vergence eye movement tests, and using contextual information (e.g. target signals and latency of the eye movement relative to the target and confidence level of the quality of eye movement recording) to improve accuracy while reducing noise interference. Additionally, we introduce a novel hybrid architecture that combines the weight-sharing feature of convolution layers with the long-range capabilities of the transformer architecture to improve model efficiency and reduce the computation cost by a factor of 3.36, while still being competitive in terms of macro F1 score. Evaluated on two diverse datasets, our method demonstrates promising results, the most powerful discrimination being Attention & Neurologic; with a macro F1 score of up to 78.8%; disorder. The results indicate the effectiveness of our approach in classifying eye movement data from different pathologies and different clinical centers accurately, thus enabling the creation of an assistant tool in the future.

Can GPT-3.5 generate and code discharge summaries?

The aim of this study was to investigate GPT-3.5 in generating and coding medical documents with International Classification of Diseases (ICD)-10 codes for data augmentation on low-resource labels. Employing GPT-3.5 we generated and coded 9606 discharge summaries based on lists of ICD-10 code descriptions of patients with infrequent (or generation) codes within the MIMIC-IV dataset. Combined with the baseline training set, this formed an augmented training set. Neural coding models were trained on baseline and augmented data and evaluated on an MIMIC-IV test set. We report micro- and macro-F1 scores on the full codeset, generation codes, and their families. Weak Hierarchical Confusion Matrices determined within-family and outside-of-family coding errors in the latter codesets. The coding performance of GPT-3.5 was evaluated on prompt-guided self-generated data and real MIMIC-IV data. Clinicians evaluated the clinical acceptability of the generated documents. Data augmentation results in slightly lower overall model performance but improves performance for the generation candidate codes and their families, including 1 absent from the baseline training data. Augmented models display lower out-of-family error rates. GPT-3.5 identifies ICD-10 codes by their prompted descriptions but underperforms on real data. Evaluators highlight the correctness of generated concepts while suffering in variety, supporting information, and narrative. While GPT-3.5 alone given our prompt setting is unsuitable for ICD-10 coding, it supports data augmentation for training neural models. Augmentation positively affects generation code families but mainly benefits codes with existing examples. Augmentation reduces out-of-family errors. Documents generated by GPT-3.5 state prompted concepts correctly but lack variety, and authenticity in narratives.

Detecting hate crimes through machine learning and natural language processing

ABSTRACT Misidentification and misreporting of hate crimes by victims and law enforcement are significant barriers to accurate data collection of hate crimes, and their consequent study and prevention. The use of machine learning in crime detection can improve the accuracy and speed at which reported incidents with bias elements are identified. This study develops a machine learning classifier that categorizes police reports as either events with bias elements or events with no bias elements. We use incident/offense reports from the Seattle Police Department to train a Natural Language Processing classification algorithm. We collect narratives, location data, and victim and suspect demographics to use as features. We evaluate the performance of logistic regression, random forest, and XGBoost algorithms, as well as several text embedding techniques. Despite substantial class imbalance, our model achieves a macro F1-score of 0.79, demonstrating the benefits of applied machine learning in accurately detecting and reporting hate crimes.

ConvNext Mixer‐Based Encoder Decoder Method for Nuclei Segmentation in Histopathology Images

ABSTRACTHistopathology, vital in diagnosing medical conditions, especially in cancer research, relies on analyzing histopathology images (HIs). Nuclei segmentation, a key task, involves precisely identifying cell nuclei boundaries. Manual segmentation by pathologists is time‐consuming, prompting the need for robust automated methods. Challenges in segmentation arise from HI complexities, necessitating advanced techniques. Recent advancements in deep learning, particularly Convolutional Neural Networks (CNNs), have transformed nuclei segmentation. This study emphasizes feature extraction, introducing the ConvNext Mixer‐based Encoder‐Decoder (CNM‐ED) model. Unlike traditional CNN based models, the proposed CNM‐ED model enables the extraction of spatial and long context features to address the inherent complexities of histopathology images. This method leverages a multi‐path strategy using a traditional CNN architecture as well as different paths focused on obtaining customized long context features using the ConvNext Mixer block structure that combines ConvMixer and ConvNext blocks. The fusion of these diverse features in the final segmentation output enables improved accuracy and performance, surpassing existing state‐of‐the‐art segmentation models. Moreover, our multi‐level feature extraction strategy is more effective than models using self‐attention mechanisms such as SwinUnet and TransUnet, which have been frequently used in recent years. Experimental studies were conducted using five different datasets (TNBC, MoNuSeg, CoNSeP, CPM17, and CryoNuSeg) to analyze the performance of the proposed CNM‐ED model. Comparisons were made with various CNN based models in the literature using evaluation metrics such as accuracy, AJI, macro F1 score, macro intersection over union, macro precision, and macro recall. It was observed that the proposed CNM‐ED model achieved highly successful results across all metrics. Through comparisons with state‐art‐of models from the literature, the proposed CNM‐ED model stands out as a promising advancement in nuclei segmentation, addressing the intricacies of histopathological images. The model demonstrates enhanced diagnostic capabilities and holds the potential for significant progress in medical research.