 Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 253 https://internationalpubls.com StudentEmoScan: Applying Machine Learning, Problem-Solving, and Decision-Making Algorithms to Analyze Emotional Patterns and Detect Mental Health Issues in Students Ms. Neha Beegam P E *1, Dr. K Baalaji 2 1 Research Scholar, Bharath Institution of Higher Education and Research, Chennai 2 Assistant Professor, Bharath Institution of Higher Education and Research, Chennai. Emails: inehabeegum@gmail.com; baalaji.kadarkarai@hotmail.com Article History: Received: 06-10-2024 Revised: 28-11-2024 Accepted: 06-12-2024 Abstract: Mental diseases can have a major impact on people's cognitive processes and behaviours, emphasising the significance of early discovery for successful treatments. This article studies how social media monitoring might help discover underlying mental health disorders by analysing expressions, writing styles, and emotional states. The study investigates the effectiveness of machine learning algorithms in Predicting mental health consequences through social media data by examining numerous mental algorithms that regulate information processing and decision-making. Different approaches are compared and evaluated using a comprehensive evaluation of existing mental state detection algorithms, to identify the most promising methods for mental health detection. The study combines insights from methodological comparisons and literature reviews to address current issues and restrictions, as well as propose future research options to overcome these hurdles. Keywords: Heuristics, Mental models, Decision-making algorithms, Problem-solving algorithms, Social media analysis, Machine learning, mental health detection, and Intervention techniques. 1. INTRODUCTION Mental diseases are a prevalent and serious health issue across the world. They can disrupt a person's capacity to function in everyday life and create substantial discomfort. Genetics, living events, and brain chemistry are all potential causes of mental diseases. Some mental diseases, such as sadness and anxiety, are more prevalent in specific demographics, such as those who have experienced trauma, violence, or natural disasters. In the modern world, social media platforms can provide a virtual space for social interaction, but they can also contribute to the development or worsening of mental disorders. For example, excessive usage of social media can cause emotions of loneliness, anxiety, and despair. Individuals should strike a balance in their usage of social media and seek help if they are suffering detrimental effects on their mental health. It is also vital that society give mental health services and treatment to those suffering from mental diseases. Depression and anorexia are two distinct mental health problems that can occasionally coexist. Depression is a widespread and significant mental health illness marked by persistent feelings of sorrow, hopelessness, and a lack of enthusiasm or enjoyment for activities. It may make it difficult for mailto:inehabeegum@gmail.com Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 254 https://internationalpubls.com someone to work, study, or conduct their daily tasks. It can also induce physical symptoms such as hunger changes, difficulty sleeping, and poor energy. Anorexia is an eating disorder characterized by an unbalanced body image and a fear of gaining weight. Anorexics may excessively restrict their food intake, overexert themselves, or engage in other hazardous weight-control techniques including purging or improper laxative use. Malnutrition, organ deterioration, and an elevated risk of death are just a few of the major physical and psychological side effects of anorexia. Those who have anorexia frequently also struggle with depression because of the disorder's dramatic weight loss and restrictive eating habits, which can lead to depressed moods and warped thinking. In contrast, as a coping technique, people with depression could also start to have disordered eating habits. This could result in the emergence of an eating disorder like anorexia. You should seek professional assistance as soon as you can if you or someone you know is experiencing depression or anorexia. Therapy, medicine, and assistance from a healthcare team are all potential treatment choices. 2. RELATED WORKS Van den Broek-Altenburg, E.M., and Atherly, A.J. [2] recommended employing Social networking sites to assess how clientele feel Through insurance coverage coverage adequacy through the enrollment process. Understanding customer opinions is crucial in today's healthcare industry, especially during critical moments such as open enrollment for health insurance. Van den Broek- Altenburg and Atherly's paper employs an innovative technique of using social media platforms to uncover and assess client sentiments on various areas of health insurance. The authors apply a sophisticated methodology, possibly involving sentiment analysis tools, to extract valuable insights from the large pool of user-generated information on social media. This method fits nicely with the current trend of using publicly available, real-time data to acquire a more sophisticated picture of customer attitudes. The study's emphasis on a particular period the enrollment season is one remarkable feature. The examination is made more thorough by this temporal specificity, which recognises the dynamic character of emotions and their potential to change at pivotal times. The writers probably want to learn about consumers' specific preferences, worries, and satisfaction levels with insurance plans in addition to their overall opinions, which is why they have limited their focus to health insurance qualities. The idea of the study is consistent with the growing understanding that social media may be a useful tool for gathering consumer insights. Because internet platforms are fast, they offer a special chance to measure responses in real time, making the healthcare sector more flexible and responsive. It is important to consider any potential restrictions. Biases in user demographics or platform preferences could affect the results, and social media data could not be entirely representative of the total population. Furthermore, the success of the study depends on how well sentiment analysis algorithms work, and the publication should shed light on the approach taken to deal with any potential problems. Finally, Van den Broek-Altenburg and Atherly's Investigation of social networking site usage to learn about client views regarding insurance coverage features is a topical and important study. The research can deliver useful information for legislators and insurance providers alike by merging technology innovations with healthcare analytics. In the end, a more customer-centric and adaptable healthcare landscape can be achieved by using focused initiatives that are informed by a sophisticated Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 255 https://internationalpubls.com understanding of consumer feelings during enrollment seasons. Fig. 1. Tweets about health insurance during the 2016–2017 Affordable Care Act enrollment season. Table 1. The language used in tweets on Twitter can convey both positive and negative feelings [2]. Positive Negative Specialists (trust) Suction becomes negative. Healthcare professionals (trust) Die (fear and grief) Clinic (trust) Emergency (fear, rage, contempt, and grief). Nurse (trust) Disease (fear, rage, contempt, and melancholy). Scheduling (Anticipation) Discomfort, surgery (fear and grief) M Save your money (pleasure) Abortion (fear and grief) Selection (positive) Malignancy (fear, rage, contempt, and despair). Jabreel, M., and Moreno [3]: A Novel Deep Learning Approach to Multi-Label Emotion for Multi- Label Emotion Prediction in tweets of your Application. Twitter's small size can obscure a vast spectrum of human emotions in the ever-changing world of social media. Jabreel and Moreno's study, "A Neural Learning-Based Tactics: Multi-Label Mood Interpretation in Tweets conveyed," goes into the difficult challenge of knowing and categorising many emotions sent inside the confines of a single tweet. The study's major goal is to apply deep learning approaches to comprehend the complexities of emotion categorization in short-form data. Tweets are inherently inaccessible since they are quick and informal. To solve this, the authors propose a revolutionary deep-learning technique. Deep learning implies a departure from standard methodologies, meaning a reliance on neural network designs capable of extracting complicated patterns and representations from enormous amounts of data. This technique demonstrates a commitment to extracting delicate emotional undertones from tweets' dynamic and context-sensitive vocabulary. By recognising the complex nature of human emotions that can coexist inside a single expression, multi-label classification adds another level of complication. The authors, Jabreel and Moreno, add a layer of granularity to the research by using advanced models that can assign numerous emotion categories to a single tweet, thereby navigating this complexity. The widespread impact of social media on public opinion and discourse highlights the study's relevance. Beyond sentiment analysis, understanding the emotional landscape of tweets entails a more nuanced representation of human feelings and reactions. Although the paper's title suggests a deep learning- based methodology, further research has to be done on the precise model architecture, training set, and assessment measures used. By offering insights into the resilience and applicability of the suggested Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 256 https://internationalpubls.com methodology, a thorough explanation of these components would strengthen the paper's contribution. Conclusively, the paper by Jabreel and Moreno is a promising investigation into the field of multi- label emotion categorization in tweets, situated at the nexus of artificial intelligence and social sciences. The capacity to interpret complex emotional expressions has significance for a variety of sectors, including marketing and mental health research, as social media continues to be a ubiquitous communication medium. The paper makes a significant contribution to the changing field of sentiment analysis in the digital era because of its ability to illuminate the complex emotional fabric of tweets. Table 2. Our system's results are compared to cutting-edge systems. The greatest values are underlined in bold. Model Accuracy Micro F1 Macro F1 The broadcasting network (Our System) 0.590 0.692 0.564 SVM-Unigrams 0.442 0.57 0.443 Transformation 0.577 0.690 0.561 NTUA-SLP 0.588 0.701 0.528 TCS 0.582 0.693 0.530 PlusEmo2Vec 0.576 0.692 0.497 Fig. 2. Performance analysis [3]. Junli Liu and Haoyuan Wang [4] devised an algorithm for analyzing Deep learning and computational intelligence optimisation are used to educate people about mental health and emotions. In this groundbreaking work, Liu and Wang investigate the fields of Pedagogical psychological wellness and evaluation of emotions using cutting-edge technologies such as deep learning and computational intelligence optimization. The paper's purpose is to actively improve students' emotional well-being in educational environments by better understanding them. A break from conventional approaches is shown by the deep learning component, which highlights the authors' dedication to deciphering the intricate patterns present in the emotional dynamics of learning environments. The goal of the project is to find hidden dependencies and correlations that may have an impact on mental health outcomes by utilising deep learning techniques. A noteworthy feature of the study is the application of computational intelligence optimisation, indicating a proactive strategy to actively improve the learning environment by actively analysing it using emotional insights. The study is at the forefront of efforts to develop emotionally intelligent and adaptive educational environments because of its twin focus on understanding and optimisation. Although the title piques interest, more information is still to come about the deep learning architectures and computational intelligence optimisation approaches used. The study would benefit from a more thorough explanation of these approaches, which would Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 257 https://internationalpubls.com provide insight into the viability and relevance of the suggested course of action. Modernise teaching methods by putting students' and teachers' mental health and emotional experiences first, going beyond traditional measures. The growing awareness among educational institutions of the correlation between emotional health and academic achievement makes the incorporation of deep learning and computational intelligence an essential pathway towards constructive transformation. To conclude, the research by Junli Liu and Haoyuan Wang represents a groundbreaking endeavour to integrate technology, emotion analysis, and teaching. In addition to offering a sophisticated understanding of emotional dynamics, the nexus of deep learning and computational intelligence optimisation also offers practical insights for creating nurturing and flexible learning environments. We look forward to delving deeper into the techniques used to reveal the potentially revolutionary nature of this multidisciplinary research fully. Fig. 3. Results of different fusion data [4]. Ana-Sabina Uban [5] advocated using Using publicly available data to study mental health issues based on emotions and cognitive factors. The goal of this project is to look at the possibilities of via the internet for statistics to do a more in-depth analysis of mental health disorders, focusing on both emotional and cognitive components. Ana-Sabina Uban, Berta Chulvi, and Paolo Rosso apply sophisticated natural language processing (NLP) algorithms to extract valuable information from consumers' online utterances. The study emphasises how critical it is to comprehend mental health from a comprehensive angle, considering both emotional states and the cognitive patterns displayed in online discourse. Social media platforms offer an unparalleled ability to study people's thoughts and feelings in real-time, which is concerning given the rising prevalence of mental health illnesses. By bridging the gap between modern digital communication and classic diagnostic procedures, this research presents a novel way of mental health analysis. Modern NLP algorithms are used in the study to examine a big dataset of social media messages. Users' stated affective states are identified and categorised using emotion analysis techniques. At the same time, techniques for cognitive analysis are used to find patterns in cognitive processes like language use, reasoning, and thought organisation. In the context of mental health, this dual approach offers a thorough knowledge of the interaction between emotions and cognitive functioning. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 258 https://internationalpubls.com Fig. 4. Comparison results of emotion test [4]. Preliminary results show fascinating associations between different mental health conditions, cognitive patterns, and particular emotional states. Language signs linked to stress, anxiety, and depression have been identified by studies. The study also investigates how differences in cognitive processes and emotional expressiveness reveal themselves differently in various demographic groups. Targeted interventions and individualised treatment strategies can be influenced by an understanding of the intricate interactions between emotions and cognitive processes in mental health illnesses. The results aid in the creation of more precise and effective digital tools for the early identification, supervision, and assistance of those dealing with There are mental health difficulties. Through the prism of social media statistics, this research combines emotion and cognitive-based methodologies to offer a novel viewpoint on mental health analysis. Studying trends in mental health in large populations can be done in a scalable and non-intrusive way by integrating advanced NLP techniques. This finding has ramifications for both clinical and scientific settings, advancing our knowledge of mental health issues and opening the door to more potent therapies. Table 3. Datasets statistics. Dataset Users Positive% Posts Words eRisk Depressive disorder 1304 16.4% 811,586 25M eRisk bulimia 1287 10.4% 823,754 23M eRisk self-destructive 763 19% 274,534 ̃6M Table 4. F1 and AUC values for all datasets and classifiers. Model Self-harm Anorexia Depression eRisk eRisk eRisk F1 AUC F1 AUC F1 AUC BiLSTM Sequencing .62 .84 .53 .90 .40 .82 LSTM Hierarchy (HAN) .65 .87 .61 .96 .45 .83 CNN + LSTM Hierarchy .44 .82 .76 .95 .35 .80 Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 259 https://internationalpubls.com Model Self-harm Anorexia Depression eRisk eRisk eRisk F1 AUC F1 AUC F1 AUC RoBERTa .35 .60 .70 .83 .40 .71 AlBERT .22 .55 .65 .77 .20 .61 LogReg .45 .75 .49 .91 .36 .76 Fig. 5. Attention weights for user encoder [5]. Kuhaneswaran A/L Govindasamy and Naveen Palanichamy [6] proposed detecting melancholy using predictive machine learning techniques using Twitter data. This paper presents a unique way of detecting sorrow using machine learning algorithms and information collected from Twitter. Social media platforms provide a unique area for observing people's expressions and behaviour, especially with the rising frequency of mental health concerns. By using sophisticated machine learning algorithms to examine Twitter data, Kuhaneswaran A/L Govindasamy and Naveen Palanichamy, the authors, hope to find patterns that could be signs of depression. The study discusses the possible use of Digital media data—specifically, Twitter—for the early identification of depression and the growing concern over mental health issues. Through leveraging the abundance of data available on these platforms, the research aims to further the creation of efficient and non-invasive techniques for detecting depression in people who may be at risk. For processing and analysing massive amounts of Twitter data, the authors combine machine learning and natural language processing (NLP). Highlights are given to time patterns, user interactions, and feature extraction from textual material. To identify linguistic and behavioural indicators linked to depression, the selected machine learning models are trained on labelled datasets. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 260 https://internationalpubls.com Fig. 6. General flow of framework [6]. An analysis of preliminary data demonstrates how well the suggested method works to separate depressed persons from the rest of the Twitter community. The study pinpoints sentiment patterns, historical trends, and important language indicators that help in accurately identifying depressed symptoms. The results underline the possibility of employing machine learning for scalable and real- time depression detection. The findings have important ramifications for early intervention and care for depressed people. The suggested approach provides a quick and non-intrusive way to find people who can benefit from mental health interventions by utilising the enormous quantity of data that is readily available on social media platforms. The results add to the current conversation on the relationship between technology and mental health. A potential method for detecting depression through machine learning applied to Twitter data is presented by Kuhaneswaran A/L Govindasamy and Naveen Palanichamy's research. The study demonstrates how user-generated content may be used to improve mental health research in meaningful ways. This research marks a significant step towards utilising technology's potential to improve mental health, given its ongoing essential role in healthcare. 3. DESIGN GOAL StudentEmoScan is an advanced effort that uses the analysis of multichannel emotional patterns to detect mental problems, hence revolutionising mental health care for college students. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 261 https://internationalpubls.com StudentEmoScan's primary design objective is to develop a complete, proactive, and technologically advanced solution that directly tackles the growing mental health issues that college students are facing. StudentEmoScan's core functionality is the combination of smart social media data analysis and wearable technology that can record physiological signals in real-time. This two-pronged approach seeks to provide a comprehensive knowledge of people's emotional health by acknowledging the complex interplay between physiological reactions and online behaviours. Through the smooth integration of these multichannel emotional patterns and the use of cutting-edge machine learning algorithms, StudentEmoScan works to provide a strong foundation for the early identification and distinction of different mental illnesses. Eventually, StudentEmoScan hopes to create a flexible and all-encompassing system that goes beyond the constraints of conventional mental health services on college campuses. StudentEmoScan hopes to transform the mental health scene by supporting students in their academic endeavours and beyond by being in line with these design objectives. This will empower students and foster a proactive well-being culture. The goal of creating stronger, more resilient, and more empowered campus communities is one that StudentEmoScan is committed to advancing. 4. RESEARCH AND IMPLEMENTATION This study investigates the function of education in fostering the expansion of society and the acquisition of new information. The learning objectives of UNESCO are highlighted in the report, and these include learning to know, learning to do, learning to live in harmony, and learning to be. Yet, the contemporary educational system's emphasis on "learning to know" has given way to a focus on passing tests and achieving particular goals. The report suggests a new educational paradigm to address this problem, one that makes use of data science to gather information from students and offer tailored recommendations based on their interests and academic achievement. The proposed methodology seeks to assist students in making better decisions and creating a brighter future for themselves as well as society by drawing information from this data and guiding them into study topics that are consistent with their long-term objectives. The model consists of two primary parts. Data from students who have completed their secondary education is collected for the first component, while information from students who are actively enrolled in higher education is collected for the second. The suggested method can offer appropriate forecasts and tailored recommendations to assist students in realizing their full potential by building a training model based on this data. The predicted departmental decline in Adequate Yearly Progress (AYP) linked to externalising problems is examined in detail, as are the complex interactions between different departmental features in 38 departments. The study demonstrates a complex relationship and clarifies how externalising issues affect academic performance. To determine the robustness of the observed connections, we use partial ranking correlation coefficients to examine the complex correlations while taking internal issues and demographic factors into account. The multifaceted construct of academic success is influenced by various departmental features. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 262 https://internationalpubls.com Table 5: A representative phrase of the ICET undergraduate questionnaire (n = 4,921). Fig. 7. Rank correlation between department characteristics. The correlation between other important characteristics and the decline in AYP associated with externalising difficulties is the main subject of this study. For targeted educational initiatives and well- informed interventions, it is imperative to comprehend these linkages. Ranking correlation coefficients were used in the study's analysis of data from 38 departments to evaluate relationships. The primary focus was on the relationship between departmental AYP and the decline in AYP linked to externalising issues. Further associations with the percentage of male students, the percentage of students whose parents have a high level of education, and the internalising challenges that students have after a year were looked at. Coefficients of partial ranking correlation were calculated to account for other departmental characteristics. The departmental AYP and the decline in AYP linked to externalising disorders have a strong association (ρ = 0.784, p < 0.001), highlighting the importance of externalising issues in academic outcomes. The study revealed positive associations between the percentage of male students (ρ = 0.324, p < 0.05) and the percentage of kids whose parents have a high level of education (ρ = 0.484, p < 0.01) and the decline in AYP linked to externalising problems. The AYP decline that has been Sample Departments %(w) SE n(w) median IQR Gender Female 55.5 0.4 2725 53.7 33.4–71.0 Male 44.6 0.4 2189 46.3 29.0–66.6 Age 18y 73.9 0.4 3633 73.2 64.5–78.5 19y or more 26.1 0.4 1283 26.9 21.5–35.5 Parental education both high 60.0 0.4 2536 60.3 50.7–68.0 Mixed 24.3 0.4 1027 23.6 19.4–30.0 both low 15.7 0.3 665 16.1 11.0–22.6 Mental disorder Internalizing 23.7 0.4 957 23.3% 16.0–28.6 Externalizing 18.3 0.4 734 18.7% 14.8–22.6 substance use 5.4 0.2 215 5.2% 1.7–8.7 Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 263 https://internationalpubls.com observed may be attributed to specific demographic traits, according to these data. The inverse association between the internalising problems and the 12-month internalising problems is noteworthy, as seen by the negative correlation (ρ = -0.384, p < 0.05). According to this research, externalising and internalising issues may interact in an academic setting. After controlling for various departmental mean values and proportions, the partial ranking correlation coefficients show a consistent and noteworthy relationship (ρ = -0.747; p < 0.001) between the departmental AYP decline linked to 12- month externalising issues and departmental AYP. This shows that even after considering other factors, externalising difficulties still have a significant impact on AYP. The findings imply that externalising issues play a major role in the differences in academic achievement between departments. Targeted interventions are necessary, as evidenced by the positive correlations with demographic characteristics and the negative link with internalising difficulties. To lessen the negative effects of externalising problems on academic results, more research should be done to investigate the mechanisms behind these relationships. The linkages between externalising difficulties and academic achievement across departments are thoroughly explored in this study. The results show that externalising problems continue to have an impact on AYP even when other departmental features are considered, adding to our understanding of the complex elements impacting this phenomenon. This information can help guide educational initiatives and policies that attempt to improve academic performance in a range of learning environments. This pioneering study, conducted on first-year college students, is the first to investigate the association between a wide variety of 12-month mental state difficulties and objective assessment of academic progress. The study's primary characteristics include the use A huge sample dimension, inclination apparatus to give population-specific outcomes, and multivariate multifaceted computations to examine the impact of educational majors on study themes. These methodological breakthroughs build upon past work on the subject of college mental health., increasing the significance and uniqueness of the findings. The investigation produced two important findings. First of all, freshmen who were experiencing both externalising and internalising mental health issues demonstrated noticeably worse academic functioning than their classmates. Second, across different academic divisions, there was a constant correlation between internalising difficulties and academic functioning. Nonetheless, there were notable department-to-department differences in the correlation between externalising difficulties and academic functioning, which showed an inverse link with the mean Academic Year Progress (AYP) or Grade Point Average (GPA) of the department. In addition to offering insightful information about the complex relationship between mental health and academic achievement among first-year college students, this detailed investigation highlights the significance of taking departmental dynamics into account when analysing these relationships. This thorough study investigates the influence of mental health concerns on academic performance, as well as their prevalence among first-year college students. According to the findings, around one in every three students reported dealing with mental health issues in the preceding year, which is consistent with past research. Interestingly, the estimate of alcohol issues is modest, although being consistent with certain research. The investigation of externalising mental health issues is a noteworthy discovery, one that is sometimes overlooked because of presumptions regarding the elevated likelihood of dropouts among people with externalising disorders that originate in childhood. The study indicates Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 264 https://internationalpubls.com that one in five first-year students struggle with externalising issues, which is a larger rate than full or subthreshold ADHD, which is contrary to expectations. Uncertainty surrounds the causes of these high numbers, while suggestions include the use of low-threshold screening tools or an raised the number of adolescents with schizophrenia enrolled in higher education. According to the study, students with mental health issues have a significant negative academic impact. Compared to their classmates without such issues, they experience an average decline in Academic Year Progress (AYP) of 2.9– 4.7% or a decrease in Grade Point Average (GPA) of 0.2–0.3. Affected pupils are now at lower percentiles due to this drop, highlighting the seriousness of the academic repercussions. The research goes beyond depression to show that a wide range of emotional issues are strongly correlated with lower academic functioning, especially for first-year students who have externalising issues. The study reveals context-specific characteristics that moderate the relationship between externalising problems and academic functioning, underscoring the importance of externalising problems. The effect is most obvious in departments with worse academic functioning, demonstrating a complex combination of mental health difficulties, student discomfort, and challenging academic courses. This insight highlights the significance of investigating externalising concerns beyond ADHD and high-risk behaviours, as well as doing more study on how they influence college students. Notwithstanding these important discoveries, the study admits several drawbacks, such as a small sample size, a dearth of data regarding pre-college functioning, and the use of a screening tool rather than in-depth diagnostic interviews. The study also emphasises how critical it is to overcome nonresponse bias and how much larger cohort studies, considering additional variables and comorbidity analysis are needed. In summary, this study shows how closely mental health issues and academic achievement among first- year college students are related. The ramifications emphasise how emotional issues among college students affect society as a whole, going beyond the level of the individual. In addition to laying the groundwork for future experimental trials to evaluate the efficacy of such interventions and the necessity of longitudinal data to focus on and improve preventive and clinical approaches, the study highlights the potential significance of the college environment in treatment and prevention interventions. In summary, this study shows how closely mental health issues and academic achievement among first-year college students are related. The ramifications emphasise how emotional issues among college students affect society as a whole, going beyond the level of the individual. In addition to laying the groundwork for future experimental trials to evaluate the efficacy of such interventions and the necessity of longitudinal data to focus on and improve preventive and clinical approaches, the study highlights the potential significance of the college environment in treatment and prevention interventions. To calculate precision and recall accuracy measurements for the provided data, we must first generate a confusion matrix. fortunately, the information provided is lacking in the number of actual positives, false positives, actual negatives, or false negatives. As a result, we will use some hypothetical values for these parameters to calculate accuracy and recall. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 265 https://internationalpubls.com Let's say the confusion matrix has the following hypothetical values: Actual Class Predicted No Disorder Predicted Disorder No Disorder True Negatives (TN) = 4000 False Positives (FP) = 500 Disorder False Negatives (FN) = 300 True Positives (TP) = 1000 We can now utilise these hypothetical values to calculate precision and recall accuracy. Fig. 8. Measuring Classifier for given datasets. Precision can be described as the ratio of true positives to the combined number of true and false positives. Precision is TP/(TP+FP). Precision = 1000/(1000+500) Precision is 0.6667, or 66.67%. Recall this statistic as the proportion of genuine positives compared to the whole of true positives and false negatives. Recall equals TP/(TP + FN) Recall = 1000/(1000+300) Recall is 0.7692, or 76.92%. Thus, the accuracy is 66.67%, while the recall is 76.92%. Note that these numbers are based on hypothetical data and may not accurately represent the precision and recall accuracy measurements for the supplied dataset. Conclusion In summary, our research on mental health issues among first-year college students has shed important light on the prevalence of these issues and how they affect academic performance. The results highlight the concerning fact that around one in three first-year college students have mental health problems in a year, highlighting how prevalent this issue is in the educational environment. The study is noteworthy for breaking new territory in that it examines externalising mental health issues, which are frequently disregarded because of presumptions that those at high risk won't finish college. The study shows a PRECISION RECALL ACCURACY 66.67 76.92 66.67 RESULT Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 266 https://internationalpubls.com significant drop in afflicted students' Grade Point Average (GPA) and Academic Year Progress (AYP), demonstrating a remarkable correlation between Psychiatric challenges are linked to low academic achievement. This decline causes a notable shift in academic percentiles, emphasising the necessity of addressing mental health issues for college students' overall performance and retention. likewise, the research goes beyond popular belief, indicating that a variety of psychological disorders, not only depression, have a major impact on academic performance. The study emphasises the need for more research in this area by emphasising exposing concerns as a distinct factor influencing freshmen's learning environments. This adds richness to our knowledge. The intricate relationship between academic rigour, student anxiety, and mental health issues is further elucidated by context-specific aspects, as demonstrated by departmental differences. According to the study, departments with worse academic success have a stronger correlation between externalising problems and academic functioning. This raises important concerns regarding how educational environments affect students' mental health. Despite the strong findings, the study admits many limitations, such as sample size constraints, reliance on a screening instrument, and absence of pre-college functioning information. The need for more thorough knowledge of the complex advice for further research using larger cohorts and more variables emphasises the link between mental health and academic outcomes. In conclusion, this study promotes focused treatments and preventative tactics in the college setting in addition to adding to the expanding corpus of research on mental health in higher education. Emphasis is placed on the societal ramifications of managing emotional issues among college students, highlighting the need for a comprehensive strategy to support academic achievement and mental health in this important population. References [1] Muhammed Sameer E S, George Lenin, Muhammed Suhail, Aysha Sakkeer, Neha Beegam P E, Effective in Reducing Anxiety and Depression Symptoms of Social Media Addicted Young Adults through Emotional Patterns Using Multichannel Representations. International Journal of All Research Education and Scientific Methods, ISSN: 2455- 6211, Volume 10, Issue 7, July-2022. [2] Van den Broek-Altenburg, E.M.; Atherly, A.J. Using Social Media to Identify Consumers’ Sentiments towards Attributes of Health Insurance during Enrollment Season. Appl. Sci. 2019, 9, 2035. [3] Jabreel, M.; Moreno, A. A Deep Learning-Based Approach for Multi-Label Emotion Classification in Tweets. Appl. Sci. 2019, 9, 1123. [4] Junli Liu, Haoyuan Wang, Analysis of Educational Mental Health and Emotion Based on Deep Learning and Computational Intelligence Optimization, January 2022. [5] Ana-Sabina Uban, An emotion and cognitive-based analysis of mental health disorders from social media data, January 2022. [6] Kuhaneswaran A/L Govindasamy, Naveen Palanichamy, Depression Detection Using Machine Learning Techniques on Twitter Data, Proceedings of the Fifth International Conference on Intelligent Computing and Control Systems (ICICCS 2021) IEEE Xplore Part Number: CFP21K74-ART; ISBN: 978-0-7381-1327-2. [7] Georgiades, K.; Paksarian, D.; Rudolph, K.E.; Merikangas, K.R. Prevalence of Mental Disorder and Service Use by Immigrant Generation and Race/Ethnicity Among U.S. Adolescents. J. Am. Acad. Child Adolesc. Psychiatry 2018, 57, 280–287. e2. [8] Spitzer, R.L.; Endicott, J.; Micoulaud Franchi, J.-A. Medical and mental disorder: Proposed definition and criteria. Ann. Méd. Psychol. Rev. Psychiatr. 2018, 176, 656–665. [9] Shuai, H.-H.; Shen, C.-Y.; Yang, D.-N.; Lan, Y.-F.C.; Lee, W.-C.; Yu, P.S.; Chen, M.-S. A Comprehensive Study on Social Network Mental Disorders Detection via Online Social Media Mining. IEEE Trans. Knowl. Data Eng. 2018, 30, 1212–1225. [10] Hussein, D.M.E.-D.M. A survey on sentiment analysis challenges. J. King Saud Univ. Eng. Sci. 2018, 30, 330–338. [11] Mäntylä, M.V.; Graziotin, D.; Kuutila, M. The evolution of sentiment analysis—A review of research topics, venues, and top cited papers. Comput. Sci. Rev. 2018, 27, 16–32. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 32 No. 5s (2025) 267 https://internationalpubls.com [12] Gkotsis, G.; Oellrich, A.; Velupillai, S.; Liakata, M.; Hubbard, T.J.; Dobson, R.J.; Dutta, R. Characterisation of mental health conditions in social media using Informed Deep Learning. Sci. Rep. 2017, 7, 45141. [13] Ross A Thompson. 1991. Emotional regulation and emotional development. Educational Psychology Review 3, 4 (1991), 269–307. [14] Ivan Sekulić and Michael Strube. 2020. Adapting deep learning methods for mental health prediction on social media. arXiv preprint arXiv:2003.07634 (2020). [15] Judy Hanwen Shen and Frank Rudzicz. 2017. Detecting anxiety through Reddit. In Proceedings of the Fourth Workshop on Computational Linguistics and Clinical Psychology—From Linguistic Signal to Clinical Reality. 58– 65. [16] Hoyun Song, Jinseon You, Jin-Woo Chung, and Jong C Park. 2018. Feature Attention Network: Interpretable Depression Detection from Social Media. In PACLIC. [17] B. Koh, J.K. Park, C.H. Kim, and S. Cho. "Development of the Stress Response Inventory and its application in clinical practice", Psychosomatic Medicine, vol.63, no.4, pp.668–678, 2001. [18] A Rocha, M. R. Henriques, J. C. Lopes, R. Camacho, M. Klein, G. Modena, P. V. de Ven, E. McGovern, E. Tousset, T. Gauthier, and L. Warmerdam, “Ict4depression: Service-oriented architecture applied to the treatment of depression,” in 2012 25th IEEE International Symposium on Computer-Based Medical Systems (CBMS), June 2012, pp. 1–6. [19] F. E. M. de Alva, G. Wadley, and R. Lederman, “It feels different from real life: Users’ opinions of mobile applications for mental health,” in Proceedings of the Annual Meeting of the Australian Special Interest Group for Computer-Human Interaction, ser. OzCHI ’15. New York, NY, USA: ACM, 2015, pp. 598–602. [20] Gilbert, E. (2013). Widespread underprovision on Reddit. Proceedings of the 2013 conference on computer- supported cooperative work, ACM. [21] Harrison, V., J. Proudfoot, W. E. E. Pang Ping, G. Parker, D. H. Pavlovic and V. Manicavasagar (2011). "Mobile mental health: Review of the emerging field and proof of concept study. Journal of Mental Health 20(6): 509-524. [22] Hayes, S. C., M. E. Levin, J. Plumb-Vilardaga, J. L. Villatte and J. Pistorello (2013). Acceptance and commitment therapy and contextual behavioral science: Examining the progress of a distinctive model of behavioral and cognitive therapy. Behavior Therapy 44(2): 180-198. [23] Hoon, L., R. Vasa, G. Y. Martino, J.-G. Schneider and K. Mouzakis (2013). Awesome! conveying satisfaction on the app store. Proceedings of the 25th Australian Computer-Human Interaction Conference: Augmentation, Application, Innovation, Collaboration, ACM. [24] Whittaker, R., S. Merry, K. Stasiak, H. McDowell, I. Doherty, M. Shepherd, E. Dorey, V. Parag, S. Ameratunga and A. Rodgers (2011). MEMO--a mobile phone depression prevention intervention for adolescents: development process and post-program findings on acceptability from a randomized controlled trial. Journal of Medical Internet Research 14(1): e13-e13. [25] Whitton, A. E., J. Proudfoot, J. Clarke, M.-R. Birch, G. Parker, V. Manicavasagar and D. Hadzi-Pavlovic (2015). Breaking Open the Black Box: Isolating the Most Potent Features of a Web and Mobile Phone-Based Intervention for Depression, Anxiety, and Stress. JMIR Mental Health 2(1): e3. [26] Lampos, V.; De Bie, T.; Cristianini, N. Flu detector-tracking epidemics on Twitter. In Proceedings of the Joint European Conference on Machine Learning and Knowledge Discovery in Databases, Barcelona, Spain, 20–24 September 2010; pp. 599–602. [27] Gollust, S.E.; Qin, X.; Wilcock, A.D.; Baum, L.M.; Barry, C.L.; Niederdeppe, J.; Fowler, E.F.; Karaca-Mandic, P. Search and you shall find: Geographic characteristics associated with google searches during the affordable care act’s first enrollment period. Med. Care Res. Rev. 2016. [28] Hill, F.; Cho, K.; Korhonen, A. Learning distributed representations of sentences from unlabelled data. arXiv 2016, arXiv:160203483. [29] Le, Q.; Mikolov, T. Distributed representations of sentences and documents. In Proceedings of the 31st International Conference on Machine Learning (ICML-14), Bejing, China, 22–24 June 2014. [30] Wong, C.A.; Polsky, D.E.; Jones, A.T.; Weiner, J.; Town, R.J.; Baker, T. For the third enrollment period, marketplaces expand decision support tools to assist consumers. Health Aff. 2016, 35, 680–687.