Deep Learning Classification Research Articles

A Hybrid Target Selection Model of Functional Safety Compliance for Autonomous Driving System

The autonomous driving system faces challenges in selecting critical targets under dense environments with limited computation resources. Existing rule-based methods struggle with complex scenarios, while learning-based approaches lack interpretability and safety. This paper proposes a hybrid target selection model combining a lightweight long short-term memory (LSTM) based deep learning classifier and rule-based methods. Key input features are identified and processed to enhance training. The LSTM model is validated for accuracy and efficiency against bidirectional LSTM (Bi-LSTM) variations. Compared to the single approach, the hybrid model integrates the LSTM classifier and rule-based methods with a synthesizer, demonstrating improved accuracy, better interpretability, and a potentially higher functional safety level. Integrated into TDA4VM, the hybrid model shows timely and complementary target selection performance in actual urban and highway tests with low computation costs, proving its theoretical value and engineering prospects.

Bilingual hate speech detection on social media: Amharic and Afaan Oromo

Due to significant increases in internet penetration and the development of smartphone technology during the preceding couple of decades, many people have started using social media as a communication platform. Social media has grown to be one of the most significant components, with several benefits. However, technology also poses a number of threats, challenges, and barriers, such as hate speech, disinformation, and fake news. Hate speech detection is one of the many ways social media platforms can be accused of not doing enough to thwart hate speech on their platform. People in Bilingual and multinational societies commonly employ a code mix in both spoken and written communication. Among these, Amharic and Afaan Oromo language speakers frequently mix the two languages when conversing and posting on social media. The majority of previous study concentrated on identifying either technological favoured language or monolingual hate speech in Ethiopian languages; however, the availability of Bilingual communication in social media hampers the propagation of hate speech via social media. In this work, a Bilingual hate speech detection for Amharic and Afaan Oromo languages were conducted using four different deep learning classifiers (CNN, BiLSTM, CNN-BiLSTM, and BiGRU) and three feature extraction (Keras word embedding, word2vec, and FastText) techniques. According to the experiment, BiLSTM with FastText feature extraction is an outperforming the other algorithm by achieving a 78.05% accuracy for Bilingual Amharic Afaan Oromo hate speech detection. The FastText feature extraction overcomes the problem of out of vocabulary (OOV). Furthermore, we are working towards including others linguistic features of the languages to detect hate speech and make the resource available to facilitate further research in the area of Bilingual hate speech detection for other under-resourced Ethiopian languages.

A Deep-Learning Model for Multi-class Audio Classification of Vocal Fold Pathologies in Office Stroboscopy.

To develop and validate a deep-learning classifier trained on voice data extracted from videolaryngostroboscopy recordings, differentiating between three different vocal fold (VF) states: healthy (HVF), unilateral paralysis (UVFP), and VF lesions, including benign and malignant pathologies. Patients with UVFP (n = 105), VF lesions (n = 63), and HVF (n = 41) were retrospectively identified. Voice samples were extracted from stroboscopic videos (Pentax Laryngeal Strobe Model 9400), including sustained /i/ phonation, pitch glide, and /i/ sniff task. Extracted audio files were converted into Mel-spectrograms. Voice samples were independently divided into training (80%), validation (10%), and test (10%) by patient. Pretrained ResNet18 models were trained to classify (1) HVF and pathological VF (lesions and UVFP), and (2) HVF, UVFP, and VF lesions. Both classifiers were further validated on an external dataset consisting of 12 UVFP, 13 VF lesions, and 15 HVF patients. Model performances were evaluated by accuracy and F1-score. When evaluated on a hold-out test set, the binary classifier demonstrated stronger performance compared to the multi-class classifier (accuracy 83% vs. 40%; F1-score 0.90 vs. 0.36). When evaluated on an external dataset, the binary classifier achieved an accuracy of 63% and F1-score of 0.48, compared to 35% and 0.25 for the multi-class classifier. Deep-learning classifiers differentiating HVF, UVFP, and VF lesions were developed using voice data from stroboscopic videos. Although healthy and pathological voice were differentiated with moderate accuracy, multi-class classification lowered model performance. The model performed poorly on an external dataset. Voice captured in stroboscopic videos may have limited diagnostic value, though further studies are needed. 4 Laryngoscope, 2025.

Deep Learning to Distinguish Edema Secondary to Retinal Vein Occlusion and Diabetic Macular Edema: A Multimodal Approach Using OCT and Infrared Imaging

Background: Retinal diseases are emerging as a significant health concern, making early detection and prompt treatment crucial to prevent visual impairment. Optical coherence tomography (OCT) is the preferred imaging modality for non-invasive diagnosis. Both diabetic macular edema (DME) and macular edema secondary to retinal vein occlusion (RVO) present an increase in retinal thickness, posing etiologic diagnostic challenges for non-specialists in retinal diseases. The lack of research on deep learning classification of macular edema secondary to RVO using OCT images motivated us to propose a convolutional neural network model for this task. Methods: The VGG-19 network was fine-tuned with a public dataset to classify OCT images. This network was then used to develop three models: unimodal—the input is only the OCT B-scan; multimodal—the inputs are the OCT B-scan and diabetes information, and multi-image—the inputs are the OCT B-scan, the infrared image, and the diabetes information. Seven hundred sixty-six patients from ULS São João were selected, comprising 208 healthy eyes, 207 with macular edema secondary to RVO, 218 with DME, and 200 with other pathologies. The performance metrics include accuracy, precision, recall, F0.5 score, and area under the receiver operating characteristic curve (AUROC). Results: The multi-image model achieved better results, with an accuracy of 95.20%, precision of 95.43%, recall of 95.20%, F0.5-score of 95.32%, F1-score of 95.21%, and AUROC of 99.59% on the classification task between four classes. Conclusions: This study presents a novel method to distinguish macular edema secondary to RVO and DME using diabetes diagnosis, OCT, and infrared images. This research aims to provide a reliable tool for ophthalmologists, improving the accuracy and speed of diagnoses.

CT-based radiomics: A potential indicator of KRAS mutation in pulmonary adenocarcinoma.

This study aimed to validate a CT-based radiomics signature for predicting Kirsten rat sarcoma (KRAS) mutation status in lung adenocarcinoma (LADC). A total of 815 LADC patients were included. Radiomics features were extracted from non-contrast-enhanced CT (NECT) and contrast-enhanced CT (CECT) images using Pyradiomics. CT-based radiomics were combined with clinical features to distinguish KRAS mutation status. Four feature selection methods and four deep learning classifiers were employed. Data was split into 70% training and 30% test sets, with SMOTE addressing imbalance in the training set. Model performance was evaluated using AUC, accuracy, precision, F1 score, and recall. The analysis revealed that 10.4% of patients showed KRAS mutations. The study extracted 1061 radiomics features and combined them with 17 clinical features. After feature selection, two signatures were constructed using top 10, 20, and 50 features. The best performance was achieved using Multilayer Perceptron with 20 features. CECT, it showed 66% precision, 76% recall, 69% F1-score, 84% accuracy, and AUC of 93.3% and 87.4% for train and test sets, respectively. For NECT, accuracy was 85% and 82%, with AUC of 90.7% and 87.6% for train and test sets, respectively. CT-based radiomics signature is a noninvasive method that can predict KRAS mutation status of LADC when mutational profiling is unavailable.

Machine learning-guided analysis of CIGS solar cell efficiency: Deep learning classification and feature importance evaluation

Machine learning-guided analysis of CIGS solar cell efficiency: Deep learning classification and feature importance evaluation

Sailfish-cat algorithm-enhanced generative adversarial network for attack detection in internet of things-Fog network authentication

The internet of things (IoT) has emerged as a prominent and influential concept within the realm of computing. Various attack detection methods are devised for detecting attacks in IoT-Fog environment. Despite all these efforts, attack detection still remained as a challenging task due to factors such as low latency, resource constraints of IoT devices, scalability issues, and distribution complexities. All these challenges are addressed in this paper by designing an efficient attack detection technique named as sailfish- cat optimization-based generative adversarial network (SaCO-based GAN) tailored for the IoT-Fog framework. This proposed approach introduces the SaCO-based GAN for IoT-Fog attack detection utilizing deep learning and feature-based classification, validated through experiments showing superior performance metrics. Notably, the SaCO optimization technique is utilized to train the GAN. Experimental results demonstrate the efficacy of the SaCO-based GAN with a maximum recall of 92.15%, a maximum precision of 91.21%, and a maximum F-Measure of 92.16%, outperforming existing techniques in IoT-Fog attack detection. The paper recommends enhancing scalability, implementing real-time detection strategies, rigorously testing robustness against diverse attack scenarios, and integrating with existing IoT security frameworks for practical deployment.

Enhancing Colorectal Polyps Detection using Transfer Learning on DICOM Metadata

Colorectal and Rectum Cancer (CRC) presents significant global health challenges, necessitating early detection and precise diagnosis to achieve effective treatment and better patient outcomes. Transfer learning techniques have shown considerable promise, especially in cancer detection. This study presents a CRC prevention system based on a fusion of a pre-trained VGG16 model with dense layers for metadata processing. Experiments were performed using the CT Colonography dataset from The Cancer Imaging Archive (TCIA), applying preprocessing and class weighting to address class imbalance. The system was evaluated using accuracy, loss, recall, precision, F1-score, and AUC. This study investigated the impact of integrating DICOM patient metadata to enhance the proposed CRC prevention system. The findings indicate that the proposed MetaVGGNet model outperformed the standard VGG16, achieving greater accuracy (82%) and a marginally lower loss. This successful application has the potential to enhance CRC diagnosis and treatment and underscores the importance of incorporating metadata into deep learning classification systems, offering avenues for more effective and dependable diagnostic tools in CRC management.

Deep learning for skin melanoma classification using dermoscopic images in different color spaces

Skin cancer begins in the skin cells. The damage to the skin cells can cause genetic mutations that lead to uncontrolled growth and the formation of tumors. It is estimated that millions of people are diagnosed with skin cancer of different kinds each year. The earlier a skin cancer is diagnosed, the better the patient's prognosis and the lower their chance of complications. In this work, an efficient deep learning classification (EDLCS) to classify dermoscopic images is developed. The importance of color in the diagnosis of skin melanoma has caused color analysis to attract considerable attention from researchers of image-based skin melanoma analysis. Three different color spaces such as red-green-blue (RGB), hue-saturation-lightness (HIS) and LAB are investigated in this study. The obtained dermoscopic images are in RGB color space. The RGB dermoscopic images are first converted into HSV and LAB spaces to investigate the HSV and LAB color spaces for melanoma classification. Then, the color space converted image is fed to the proposed EDLCS to evaluate their performances. Results show that the proposed EDLCS provides 99.58% accuracy while using the LAB color model to classify preprocessed images while other models provide 99.17%.

Optimizing Skin Cancer Diagnosis: A Modified Ensemble Convolutional Neural Network for Classification.

Skin cancer is recognized as one of the most harmful cancers worldwide. Early detection of this cancer is an effective measure for treating the disease efficiently. Traditional skin cancer detection methods face scalability challenges and overfitting issues. To address these complexities, this study proposes a random cat swarm optimization (CSO)with an ensemble convolutional neural network (RCS-ECNN) method to categorize the different stages of skin cancer. In this study, two deep learning classifiers, deep neural network (DNN) and Keras DNN (KDNN), are utilized to identify the stages of skin cancer. In this method, an effective preprocessing phase is presented to simplify the classification process. The optimal features are selected using the feature extraction phase. Then, the GrabCut algorithm is employed to carry out the segmentation process. Also, the CSO is employed to enhance the effectiveness of the method. The HAM10000 and ISIC datasets are utilized to evaluate the RCS-ECNN method. The RCS-ECNN method achieved an accuracy of 99.56%, a recall of 99.66%, a specificity value of 99.254%, a precision value of 99.18%, and an F1-score value of 98.545%, respectively. The experimental results demonstrated that the RCS-ECNN method outperforms the existing techniques.

Chaotic gradient based optimization with fuzzy temporal optimized CNN for heart failure prediction

Heart failure is a leading cause of premature death, especially among individuals with a sedentary lifestyle. Early and accurate detection is essential to prevent the progression of this situation. However, many existing prediction systems failed to detect early and accurately, also taking more time to detect. To address these issues, we propose an advanced heart failure detection model that combines one-dimensional chaotic maps and a Gradient-Based Optimizer (GBO) called Chaotic Gradient-Based Optimizer (CGBO). This approach improves feature selection by effectively selecting the most crucial features related to the risk of heart failure. Additionally, we introduce the Fuzzy Temporal Optimized Convolutional Neural Network (FTOCNN) classifier that incorporates CGBO and fuzzy temporal rules to enhance detection accuracy. The proposed model is evaluated using the UCI heart dataset and Electronic Health Records (EHRs) and its performance is assessed through statistical measures, classification metrics, and a Wilcoxon rank-sum p-test. Furthermore, a tenfold cross-validation process ensures a comprehensive evaluation and the proposed method outperforms different Machine Learning (ML) / Deep Learning (DL) classifiers. The experimental findings reveal that CGBO significantly improves the predictive performance of the FTOCNN classifier by achieving 94% accuracy in EHR and enhances the reliability of heart failure detection compared to existing systems.

Application of a methodological framework for the development and multicenter validation of reliable artificial intelligence in embryo evaluation

BackgroundArtificial intelligence (AI) models analyzing embryo time-lapse images have been developed to predict the likelihood of pregnancy following in vitro fertilization (IVF). However, limited research exists on methods ensuring AI consistency and reliability in clinical settings during its development and validation process. We present a methodology for developing and validating an AI model across multiple datasets to demonstrate reliable performance in evaluating blastocyst-stage embryos.MethodsThis multicenter analysis utilizes time-lapse images, pregnancy outcomes, and morphologic annotations from embryos collected at 10 IVF clinics across 9 countries between 2018 and 2022. The four-step methodology for developing and evaluating the AI model include: (I) curating annotated datasets that represent the intended clinical use case; (II) developing and optimizing the AI model; (III) evaluating the AI’s performance by assessing its discriminative power and associations with pregnancy probability across variable data; and (IV) ensuring interpretability and explainability by correlating AI scores with relevant morphologic features of embryo quality. Three datasets were used: the training and validation dataset (n = 16,935 embryos), the blind test dataset (n = 1,708 embryos; 3 clinics), and the independent dataset (n = 7,445 embryos; 7 clinics) derived from previously unseen clinic cohorts.ResultsThe AI was designed as a deep learning classifier ranking embryos by score according to their likelihood of clinical pregnancy. Higher AI score brackets were associated with increased fetal heartbeat (FH) likelihood across all evaluated datasets, showing a trend of increasing odds ratios (OR). The highest OR was observed in the top G4 bracket (test dataset G4 score ≥ 7.5: OR 3.84; independent dataset G4 score ≥ 7.5: OR 4.01), while the lowest was in the G1 bracket (test dataset G1 score < 4.0: OR 0.40; independent dataset G1 score < 4.0: OR 0.45). AI score brackets G2, G3, and G4 displayed OR values above 1.0 (P < 0.05), indicating linear associations with FH likelihood. Average AI scores were consistently higher for FH-positive than for FH-negative embryos within each age subgroup. Positive correlations were also observed between AI scores and key morphologic parameters used to predict embryo quality.ConclusionsStrong AI performance across multiple datasets demonstrates the value of our four-step methodology in developing and validating the AI as a reliable adjunct to embryo evaluation.

DICOM LUT is a Key Step in Medical Image Preprocessing Towards AI Generalizability.

Image pre-processing has significant impact on performance of deep learning models in medicine; yet, there is no standardized method for DICOM pre-processing. In this study, we investigate the impact of two commonly used image preprocessing techniques, histogram equalization (HE) and values-of-interest look-up-table (VOI-LUT) transformations on the performance deep learning classifiers for chest X-rays (CXR). We generated two baseline datasets (raw pixel and standard DICOM processed) from our internal CXR dataset and then enhanced both with HE to create four distinct datasets. Four independent deep learning models for diagnosis of pneumothorax were trained and evaluated on two external datasets. Results reveal that HE enhancement significantly affects model performance, particularly in terms of generalizability. Models trained solely on HE-enhanced datasets exhibit poorer performance on external validation sets, suggesting potential overfitting and information loss. These models also exhibit shortcut learning, relying on spurious correlations in the training data for their prediction. This study highlights the importance of machine learning practitioners being aware of preprocessing techniques applied to datasets and their potential impacts on model performance, as well as need for including preprocessing information when sharing datasets. Additionally, this research underscores the necessity of using pixel values closer to clinical standards during dataset curation to improve model robustness and mitigate the risk of information loss.

Optimizing dental implant identification using deep learning leveraging artificial data

This study aims to evaluate the potential enhancement in implant classification performance achieved by incorporating artificially generated images of commercially available products into a deep learning process of dental implant classification using panoramic X-ray images. To supplement an existing dataset of 7,946 in vivo dental implant images, a three-dimensional scanner was employed to create implant surface models. Subsequently, implant surface models were used to generate two-dimensional X-ray images, which were compiled along with original images to create a comprehensive dataset. Images of 10 types of implants were classified using ResNet50 into the following datasets: (A) images of implants captured in vivo, (B) artificial implant images generated without background adjustments, and (C) implant images derived from in vivo images and generated with background adjustments. The classification accuracy was 0.8888 for dataset A, 0.903 for dataset B, and 0.9146 for dataset C. Notably, dataset C demonstrated the highest performance and exhibited the optimal feature distribution. In the context of deep learning classifiers for dental implants using panoramic X-ray images, incorporating artificially generated X-ray images—designed to mirror the appearance of human body implants—proved to be the most beneficial in enhancing the performance of the classification model.

Wireless, Wrist‐Worn Ultraconformal Neuromuscular Interfaces for Thumb‐Controlled Immersive Human–Machine Interactions

AbstractThumb actions, outperforming conventional methods such as hand gestures or wrist gestures in terms of dexterity, agility and intuitiveness, have long been sought‐after for achieving immersive interactive experiences in robotic control and AR/VR platforms. However, accurate mapping of such dynamic, subtle thumb actions remains a difficult challenging. Here, a wireless, wrist‐affixed soft ultra‐conformal neuromuscular interface system (UniSyst) is reported to capture high‐fidelity surface electromyography crucial for decoding dynamic subtle thumb actions. The UniSyst, which can be easily fabricated at scale and in mass quantities, features a 16‐channel soft, stretchable sensing array for broad, high‐resolution data capture and a stiff design for plug‐and‐play interface with external rigid wireless acquisition module. This soft‐stiff design ensures consistent electrode‐skin contact under substantial skin deformations, thus avoiding undesired motion artifacts commonly observed with rigid alternatives. Facilitated with a lightweight 1D convolution neural network deep learning classifier, this system shows remarkable recognition accuracy over that of traditional forearm placements, fairly compared through concurrently collected signals from the wrist and the forearm of 12 participants. In practical scenarios, the soft UniSyst exhibits rapid, precise thumb‐controlled interactive capabilities, adeptly managing human–machine communications in both digital platforms and immersive gaming controls.

Dual-hybrid intrusion detection system to detect False Data Injection in smart grids.

Modernizing power systems into smart grids has introduced numerous benefits, including enhanced efficiency, reliability, and integration of renewable energy sources. However, this advancement has also increased vulnerability to cyber threats, particularly False Data Injection Attacks (FDIAs). Traditional Intrusion Detection Systems (IDS) often fall short in identifying sophisticated FDIAs due to their reliance on predefined rules and signatures. This paper addresses this gap by proposing a novel IDS that utilizes hybrid feature selection and deep learning classifiers to detect FDIAs in smart grids. The main objective is to enhance the accuracy and robustness of IDS in smart grids. The proposed methodology combines Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO) for hybrid feature selection, ensuring the selection of the most relevant features for detecting FDIAs. Additionally, the IDS employs a hybrid deep learning classifier that integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to capture the smart grid data's spatial and temporal features. The dataset used for evaluation, the Industrial Control System (ICS) Cyber Attack Dataset (Power System Dataset) consists of various FDIA scenarios simulated in a smart grid environment. Experimental results demonstrate that the proposed IDS framework significantly outperforms traditional methods. The hybrid feature selection effectively reduces the dimensionality of the dataset, improving computational efficiency and detection performance. The hybrid deep learning classifier performs better in key metrics, including accuracy, recall, precision, and F-measure. Precisely, the proposed approach attains higher accuracy by accurately identifying true positives and minimizing false negatives, ensuring the reliable operation of smart grids. Recall is enhanced by capturing critical features relevant to all attack types, while precision is improved by reducing false positives, leading to fewer unnecessary interventions. The F-measure balances recall and precision, indicating a robust and reliable detection system. This study presents a practical dual-hybrid IDS framework for detecting FDIAs in smart grids, addressing the limitations of existing IDS techniques. Future research should focus on integrating real-world smart grid data for validation, developing adaptive learning mechanisms, exploring other bio-inspired optimization algorithms, and addressing real-time processing and scalability challenges in large-scale deployments.

Evaluating the Effect of Surrogate Data Generation on Healthcare Data Assessment

In healthcare applications, often it is not possible to record sufficient data as required for deep learning or data-driven classification and feature detection systems due to the patient condition, various clinical or experimental limitations, or time constraints. On the other hand, data imbalance invalidates many of the test results crucial for clinical approvals. Generating synthetic (artificial or dummy) data has become a potential solution to address this issue. Such data should possess adequate information, properties, and characteristics to mimic the real-world data recorded in natural circumstances. Several methods have been proposed for this purpose, and results often show that adding surrogates improves the decision-making accuracy. This article evaluates the most recent surrogate data generation and data synthesis methods to investigate the effects of the number of surrogates on improving the classification results. It is shown that the data analysis/classification results improve with an increasing number of surrogates, but this no longer continues after a certain number of surrogates. This achievement helps in deciding on the number of surrogates for each strategy, resulting in the alleviation of the computation cost.

Detecting autism in children through drawing characteristics using the visual-motor integration test.

This study introduces a novel classification method to distinguish children with autism from typically developing children. We recruited 50 school-age children in Taiwan, including 44 boys and 6 girls aged 6 to 12 years, and asked them to draw patterns from a visual-motor integration test to collect data and train deep learning classification models. Ensemble learning was adopted to significantly improve the classification accuracy to 0.934. Moreover, we identified five patterns that most effectively differentiate the drawing performance between children with and without ASD. From these five patterns we found that children with ASD had difficulty producing patterns that include circles and spatial relationships. These results align with previous findings in the field of visual-motor perceptions of individuals with autism. Our results offer a potential cross-cultural tool to detect autism, which can further promote early detection and intervention of autism.

Deep learning classification of MGMT status of glioblastomas using multiparametric MRI with a novel domain knowledge augmented mask fusion approach

We aimed to build a robust classifier for the MGMT methylation status of glioblastoma in multiparametric MRI. We focused on multi-habitat deep image descriptors as our basic focus. A subset of the BRATS 2021 MGMT methylation dataset containing both MGMT class labels and segmentation masks was used. A comprehensive mask fusion approach was developed to select relevant image crops of diseased tissue. These fusion masks, which were guided by multiple sequences, helped collect information from the regions that seem disease-free to radiologists in standard MRI sequences while harboring pathology. Integrating the information in different MRI sequences and leveraging the high entropic capacity of deep neural networks, we built a 3D ROI-based custom CNN classifier for the automatic prediction of MGMT methylation status of glioblastoma in multi-parametric MRI. Single sequence-based classifiers reached intermediate predictive performance with 0.65, 0.71, 0.77, and 0.82 accuracy for T1W, T2W, T1 contrast-enhanced, and FLAIR sequences, respectively. The multiparametric classifier using T1 contrast-enhanced and FLAIR images reached 0.88 accuracy. The accuracy of the four-input model that used all sequences was 0.81. The best model reached 0.90 ROC AUC value. Integrating human knowledge in the form of relevant target selection was a useful approach in MGMT methylation status prediction in MRI. Exploration of means to integrate radiology knowledge into the models and achieve human-machine collaboration may help to develop better models. MGMT methylation status of glioblastoma is an important prognostic marker and is also important for treatment decisions. The preoperative non-invasive predictive ability and the explanation tools of the developed model may help clinicians to better understand imaging phenotypes of MGMT methylation status of glial tumors.

A Natural Language Processing Model for Early Detection of Suicidal Ideation in Textual Data

According to the World Health Organization, suicide is one of the top 10 causes of death. An estimated 138 people's lives are significantly impacted for every suicide death, and practically every other statistic pertaining to suicide fatalities is concerning. The widespread usage of social media and the almost universal mobile devices used to access social media networks present new opportunities for preventative intervention as well as new forms of data for studying the behavior of those who (attempt to) take their own life. We show that it is feasible to identify suicidal risk individuals using social media data. Specifically, we propose concepts for an automated system for detecting measurable signals around suicide attempts using natural language processing and machine learning (particularly deep learning) approaches. The goal of this project is to improve the automatic identification and reporting of suicidal posts. It offers a method that examines Twitter as a social media platform to find warning indicators for suicide in people. The previously mentioned approach's main goal is to automatically detect anomalous changes in a user's online behavior. The comprehension and identification of intricate factors of risk or warning indicators that might precede the incident provide difficulties in the prevention of suicide. Many natural language processing (NLP) approaches are used to measure textual variations and pass them via a unique framework that may be utilized broadly to accomplish this objective. Deep learning and machine learning-based categorization algorithms are used to identify suicidal thoughts in the early stages by analyzing tweets on social networking platform Twitter. We first performed data pre-processing for both classifiers, followed by feature extraction, and then machine learning and deep learning classifiers, respectively. We use a CNN-LSTM model for this purpose to assess and contrast it with other classification methods. In comparison to earlier CNN-LSTM systems, the study demonstrates that the CNN-LSTM framework using embedding of words techniques achieves 94% classification accuracy.