|Year : 2022 | Volume
| Issue : 2 | Page : 75-84
Comparison of the performance of machine learning-based algorithms for predicting depression and anxiety among University Students in Bangladesh: A result of the first wave of the COVID-19 pandemic
Md Iqbal Hossain Nayan1, M Sheikh Giash Uddin2, Md Ismail Hossain2, Md Mohibul Alam3, Maliha Afroj Zinnia4, Iqramul Haq5, Md Moshiur Rahman6, Rejwana Ria4, Md Injamul Haq Methun7
1 Quality Services and Compliance, Square Pharmaceutical Limited, Dhaka, Bangladesh
2 Department of Statistics, Jagannath University, Dhaka, Bangladesh
3 Department of Training, Eskayef Pharmaceuticals Limited, Dhaka, Bangladesh
4 Department of Pharmacy, East West University, Dhaka, Bangladesh
5 Department of Agricultural Statistics, Sher-e-Bangla Agricultural University, Dhaka, Bangladesh
6 Department of Pharmacology and Toxicology, Sher-e-Bangla Agricultural University, Dhaka, Bangladesh
7 Department of Statistics, Tejgaon College, Dhaka, Bangladesh
|Date of Submission||04-Jan-2022|
|Date of Decision||12-Apr-2022|
|Date of Acceptance||01-May-2022|
|Date of Web Publication||23-May-2022|
Department of Agricultural Statistics, Sher-e-Bangla Agricultural University, Dhaka 1207
Source of Support: None, Conflict of Interest: None
Introduction: The purpose of this research was to predict mental illness among university students using various machine learning (ML) algorithms. Methods: A structured questionnaire-based online survey was conducted on 2121 university students (private and public) living in Bangladesh. After obtaining informed consent, the participants completed a web-based survey examining sociodemographic variables and behavioral tests (including the Patient Health Questionnaire (PHQ-9) scale and the Generalized Anxiety Disorder Assessment-7 scale). This study applied six well-known ML algorithms, namely logistic regression, random forest (RF), support vector machine (SVM), linear discriminate analysis, K-nearest neighbors, Naïve Bayes, and which were used to predict mental illness among university students from Dhaka city in Bangladesh. Results: Of the 2121 eligible respondents, 45% were male and 55% were female, and approximately 76.9% were 21–25 years old. The prevalence of severe depression and severe anxiety was higher for women than for men. Based on various performance parameters, the results of the accuracy assessment showed that RF outperformed other models for the prediction of depression (89% accuracy), while SVM provided the best result than other models for the prediction of anxiety (91.49% accuracy). Conclusion: Based on these findings, we recommend that the RF algorithm and the SVM algorithm were more moderate than any other ML algorithm used in this study to predict the mental health status of university students in Bangladesh (depression and anxiety, respectively). Finally, this study proposes to apply RF and SVM classification when the prediction of mental illness status is the core interest.
Keywords: Anxiety, Bangladesh, COVID-19, depression, machine learning algorithm, psychological
|How to cite this article:|
Nayan MI, Uddin M S, Hossain MI, Alam MM, Zinnia MA, Haq I, Rahman MM, Ria R, Haq Methun MI. Comparison of the performance of machine learning-based algorithms for predicting depression and anxiety among University Students in Bangladesh: A result of the first wave of the COVID-19 pandemic. Asian J Soc Health Behav 2022;5:75-84
|How to cite this URL:|
Nayan MI, Uddin M S, Hossain MI, Alam MM, Zinnia MA, Haq I, Rahman MM, Ria R, Haq Methun MI. Comparison of the performance of machine learning-based algorithms for predicting depression and anxiety among University Students in Bangladesh: A result of the first wave of the COVID-19 pandemic. Asian J Soc Health Behav [serial online] 2022 [cited 2022 Jun 30];5:75-84. Available from: http://www.healthandbehavior.com/text.asp?2022/5/2/75/345816
| Introduction|| |
COVID-19 started as a local transmission from the Wuhan city of China and has become one of the major calamities of the century., From the earlier status of a global health emergency, the WHO officially certified COVID-19 as a “pandemic” on February 11, 2020 (WHO, 2020). Societies are facing great uncertainty considering the knowledge being developed about the unpredictable nature of the spread of this virus and its reciprocation with societal responses.,
Mental illness is also increasing at an epidemic rate worldwide, which was severe due to fear of COVID-19., Some recent studies have shown that sociodemographic, behavior, and education are the main influencing factors for mental illness, including gender, residence, relationship status, socioeconomic status, and loneliness, personal autonomy, and future plans.,,,, From a mental health standpoint, due to COVID-19, the world population have impacted not only by anxiety and trauma but also from unfavorable societal dynamics. Taking into account the current COVID-19 pandemic, numerous universities across the world have either held over or canceled all their campus activities. Universities have shifted their programs from the face-to-face to online system., During quarantined and outside from university environment and schedule, students may encounter stress, anxiety, anger, boredom, loneliness and other emotions, which have both shorter and longer-term psychological impacts on health.,, It affects the student's energy level, concentration, reliability, mental ability and optimism, thereby affecting student performance.,
The Government of Bangladesh declared a public holiday on March 26, 2020, to lessen the transmission of COVID-19. Since then, all the schools, colleges, and universities were closed, hampering the students' studies, daily routines, and daily habits, which in turn affecting their mental health. Furthermore, home quarantine, maintaining physical distance, and other restrictions also psychologically affected students and hindered their mental well-being., On top of that, unpredictable situation, news, rumors, and misinformation can also raise negative thoughts within university students about their future., All of this together can bring hopelessness, fear of death, and frustration among quarantined students.
Before COVID-19, mental health studies reported an elevated level of moderate to extremely severe depression (52.2%), anxiety (58.1%), and stress (24.9%) among university students in Bangladesh. Yet very little information is available on the mental situation of students during the COVID-19 pandemic, and so acquiring structured measures of depression, anxiety, stress can help to estimate the necessity for interventions to diminish the mental health impacts of the pandemic on students.
Mental health is an indicator of a person's emotional, psychological, and social well-being. There are many causes of mental illness (depression and anxiety). Academic performance, occupational status, and family status are considered the most important factors leading to depression and anxiety. Furthermore, previous investigations found that different variables, for example, sex, age, marital status, were significant factors related to depression and anxiety status.,,, In addition to proper diagnosis and intervention with mental health reduces the risk for depression and anxiety. Various statistical methods have been used to evaluate the mental health status of university students. The main objective of an estimation procedure is the correct prediction of depression and anxiety status using a machine learning (ML) algorithm. ML, a scientific method that intersects artificial intelligence and statistical learning research, may be how much knowledge can be researched to look for unknown associations or trends. ML algorithms that can build models for prediction purposes have shown excellence in taking care of classification problems, in comparison to the classical statistical model. Furthermore, within the field of health and medical research, ML has become common.
Many researchers have used ML algorithms such as random forest (RF) trees, support vector machines (SVMs), and convolutional neural networks to predict anxiety and depression. In Sau et al.'s study, different ML algorithms such as logistic regression (LR), Catboost, Naïve Bayes (NB), RF, and SVM were used for classification. In addition, ML can be used to better predict the risk of depression and anxiety. Therefore, the purpose of this study is to compare the prediction performance of six well-known ML algorithms such as LR, RF, SVM, linear discriminate analysis (LDA), K-nearest neighbors (KNN), and NB have been applied to predict mental health status (depression and anxiety) among university students in Bangladesh.
| Methods|| |
Participants and procedure
The research was a prospective cross-sectional survey, and the data was collected through an online questionnaire survey using a Google form. This self-administered rapid online survey was conducted between August 2020 and September 2020. Both public and private university students operating in Dhaka city and those who have an internet facility were able to respond to the questionnaire and were eligible for this analysis. To fulfill the objective, a structured questionnaire was designed by the authors, and the questionnaire was arranged into a “Google Form” by collecting highly pertinent facts about mental health and then it was sent to various public and private university students through social media and asked them to convey their informative responses.
A multistage sampling technique was applied for this study. In the first stage, 4 universities (two public and two private universities, respectively) were chosen randomly from Dhaka city. In the second stage, six departments were chosen from each of the selected universities. In the third stage, 100 students were randomly chosen from each of the selected departments from the 1st year to the 5th year (additionally, graduate students for some departments). Here students were numbered consecutively each year as per their ID numbers. From this design, a total of 2400 students were selected for an interview and out of this 2350 consented. From the 2350 respondents, a total of 2121 students were found who had answered the structured questionnaire through “Google Form” and valuable information was stored precisely. In this case, the study sample size was 2121 and the response rate was 90.25%. Almost 10% of the students did not participate in this study after knowing the consent information. Among the students, 55.1% were women and 44.9% were men. In this analysis, the students came from all over Bangladesh and could be representative of the entire population.
The Research Ethics Committee of the Department of Agricultural Statistics at Sher-e-Bangla Agricultural University, Dhaka-1207, allowed ethical discussions to conduct the present study. All participants are informed about the purpose of the study, and the unanimity of their identity is insured, and consent from all is obtained.
Variables and measures
As the main response variable, the main concern of this study was the state of depression status and anxiety status among participants, and the two most popular methods, the Patient Health Questionnaire (PHQ-9) and Generalized Anxiety Disorder Assessment (GAD-7), were used to measure depressive symptoms and anxiety symptoms, respectively.
Patient Health Questionnaire
The Patient Health Questionnaire (PHQ-9) was used to screen the presence of depression through nine self-administered questionnaires, and every item of this questionnaire is rated on a 4-point Likert scale ranging from 0 (not at all) to 3 (nearly every day) queries the existence and rate of repetition of depressive manifestation experienced by the respondent in the last 14 days., The total score of PHQ-9 ranges from 0 to 27, and the recommended severity cut-off scores are: None (<5), mild (5–9), moderate (10–14), moderately severe (15–19), and severe (>19). The internal consistency of this scale was very high in the present study (Cronbach coefficient Alpha = 0.813).
Generalized Anxiety Disorder Assessment
In epidemiological surveys, GAD-7 was a more valid and reliable tool, which was a questionnaire consisting of 7 items and all of them carried a point based on a four-point Likert scale ranging from 0 (not at all) to 3 (Nearly every day)., It is used to screen anxiety level and assesses its severity. According to scores of 0–4, 5–9, 10–14, and 15–21, anxiety levels are divided into four categories: Mild, mild, moderate, and severe. In the present study, Cronbach's alpha was 0.878.
As independent variables, we used a set of socioeconomic and demographic factors, and all of them related to depressive symptoms were considered covariates. The factors are listed below.
- Gender (male, female)
- Current age (<20, 21–25, >25)
- Current living status (with family, without family)
- Occupation status (student, job holder, unemployed)
- Educational year (1st/2nd, 3rd/4th/5th and graduate)
- Family status (higher class, middle class, lower class)
- Marital status (married, unmarried)
- Cumulative grade point average (CGPA) (≤3.00, 3.01–3.50, >3.50).
The relationship between explanatory variables and depression status was tested in a bivariate setting. In the bivariate setting, we applied the Chi-square independence test. Mathematically, the Chi-square statistic can be defined as follows:
where Oi and Ei are the observed and expected frequency, respectively. Statistics follow the Chi-square distribution with the degrees of freedom ([number of row – 1] × [number of column – 1]).
Machine learning algorithms
We used six different supervised algorithms to predict the level of depression and anxiety among university students.
LR was a “statistical learning” technique, which is a “supervised” ML method specifically used for “classification” tasks. It uses the maximum likelihood estimation procedure to estimate the parameters of interest. Let Xi1,Xi2 …Xip be a set of explanatory variables, which can be quantitative variables or index variables that refer to the level of categorical variables, and Y is a binary variable, which has a Bernoulli distribution of a parameter πi, then the logit regression model is,
be the parameters.
RF is an ensemble learning-based classification approach with a large number of decision trees constructed in the training process, where the final output integrates the outcome class of individual decision trees.
Support vector machine
SVM is one of the most popular classification algorithms; it has a good method of transforming nonlinear data. Chen Junli and Jiao Licheng explained the classification strategy of SVM well. The linear SVM model is used in the prediction research of this mental health disease.
Linear discrimination analysis
LDA is a supervised ML technique used to extract important highlights from a data set. When these classes are well separated, the parameter estimation of the LR model is unstable. In this case, LDA is used.
The KNN algorithm is also the simplest and one of the most widely used classification algorithms. The KNN algorithm has confirmed the multiclass label classification problem and has good generalization ability. The algorithm stores each accessible case and classifies new cases based on similarity measures.
The NB classifier is a probabilistic classifier based on the assumption of strong (naive) independence between the features of Bayes' theorem. The NB model is easy to construct without estimating complex repeat parameters, which makes it particularly effective in the treatment field. Although simple, NB classifiers usually perform well and are widely used because they outperform more complex classification methods.
First, apply data preparation methods, for example, distinguish missing values from the data set and process them. Subsequently, 75% of the individual samples in each group (called the training data set) were used to apply the ML algorithm, and the remaining 25% of the individuals (called the test data set) were verified. All models were trained to support 10-fold cross-validation. We used a 10-fold cross-validation in the training set and evaluated the performance in the test set.
Six evaluation parameters were taken into account, named accuracy, sensitivity, specificity, positive predictive value, negative predictive value, Cohen's kappa. All data analyzes were completed using the Statistical Package for Social Sciences (SPSS) version 25 (IBM Corporation, Armonk, New York, NY, USA) and the R- programming (version 4.0.0, R Core Team).
| Results|| |
Socio-demographic characteristics of university students
[Table 1] lists basic demographic and psychological characteristics, such as gender, age, residence, occupation, years of education, family status, marital status, and CGPA. In this study, slightly above half of the respondents (55%) were female, and more than one-third (77%) belonged to the 21–25 years of age group. Most of the respondents live with their families, and the percentage distribution was 86.8%. The majority (approximately, 87%) of the interviewees were undergraduate students, while the rest were unemployed or employed. Nearly half of the respondents (45%) were students and studied in the 3rd year or above. Approximately two-thirds of the respondents (67%) were from household with average income (middle-income families). Almost 43% of the respondents had an average GPA between 3.01 and 3.50.
|Table 1: Percentage distribution of university students to selected sociodemographic characteristics in Bangladesh|
Click here to view
Responses to the PHQ-9 and generalized anxiety disorder-7 questionnaire
This section describes the response rates of 9 questions on the Patient Health Questionnaire and 7 questions on the General Anxiety Disorder Questionnaire. Both questionnaires are scored on a 4-point Likert scale, ranging from 0 (not at all) to 3 (almost every day). The answers are shown in [Figure 1].
|Figure 1: Percentage responses of PHQ-9 and GAD-7. GAD-7: Generalized anxiety disorder-7|
Click here to view
Association with depression and anxiety among university students
The prevalence of severe depression and severe anxiety with the background characteristics of the selected covariate is shown in [Table 2] and [Table 3], respectively.
|Table 2: Assessing association between selected covariates and depression status among university student in Bangladesh using Chi-square test|
Click here to view
|Table 3: Assessing the association between selected covariates and anxiety status among university students in Bangladesh using the Chi-square test|
Click here to view
In terms of depression level, the percentage of severely depressed students was higher, 37.1% of female students, 34.1% of respondents under the age of twenty, 27.4% of respondents living with their families, unemployed respondents (41.6%), respondents from poor families (37.2%), 1st/2nd year students (34.0%), unmarried respondents (28.0%), students in grades 3.01–3.50 (31.6%). It was found that all these selected covariates were significantly related to the depressive symptoms of university students (P < 0.001).
[Table 3] shows specifying the association between the socioeconomic and demographic characteristics and anxiety status among university students in Bangladesh. According to [Table 3], it was found that the relationship involving the student's gender and their anxiety status was profoundly noteworthy (P < 0.001). The most noteworthy rate of severe anxiety was found in female students (26.1%). There was also a significant connection between the age of the students (in years) and their anxiety status (P = 0.03). Students aged more than 25 years had severe anxiety (31.2%). Students who were currently studying (22.9%) and 1st-year/2nd-year students (29.8%) had severe anxiety. It has also shown in [Table 3] that an association between CGPA and anxiety status also exist (P < 0.001) and around 30% student with lower average result had severe anxiety level.
Performance parameter of machine learning algorithms
In this study, six different ML algorithms were used to classify the levels of depression and anxiety among university students in the test data set as severe and nonsevere. The predictive performance of these algorithms will be compared based on performance parameters such as accuracy, sensitivity, and specificity. In addition, Cohen's k statistic was used to determine the discrimination accuracy of the algorithm. [Table 4] and [Table 5], respectively show the prediction results of depression and anxiety states with performance parameters for each ML algorithm (for training and testing data sets).
|Table 4: Performance indicators of all five machine learning algorithms to predict depression status|
Click here to view
|Table 5: Performance indicators of the five machine learning algorithms to predict anxiety status|
Click here to view
The performance indicators used to predict depression are shown in [Table 4]. Using linear discriminant analysis, the accuracy in the test data set was 74.29%, the sensitivity was 22.69%, and the specificity was 93.04%. SVM (linear) showed 74.10% accuracy, 19.14% sensitivity, and 94.07% specificity in predicting the depression status of the experimental observation results. The accuracy, sensitivity, and specificity of LR were reported as 74.48%, 22.69%, and 93.29%, respectively. However, the performance shown by the k-nearest neighborhood algorithm was 88.28%, 66.67%, and 96.13%, accuracy, sensitivity and specificity, respectively. Naive Bayes showed an accuracy of 68.24%, a sensitivity of 25.53%, and a specificity of 83.76% in predicting the depression state of the test observation results. Among the six classifiers, the best result was achieved by the RF algorithm, which showed that accuracy was 88.66%, the sensitivity was 68.79%, and specificity was 95.88%.
The Cohen kappa statistics of linear discriminant analysis, SVM (linear), naive Bayes, and LR were 0.1931, 0.1664, 0.1027, and 0.1968, respectively. It was recommended to adopt a “slightly fair agreement.” However, among all executed ML algorithms, the RF algorithm showed the greatest discriminative ability (Cohen's kappa = 0.6903).
[Table 5] (performance indicators for predicting anxiety states) shows that using linear discriminant analysis, the accuracy of the test data set was 79.58%, the sensitivity was 14.87%, and the specificity was 98.75%. RF showed an accuracy of 91.30%, a sensitivity of 70.25% and a specificity of 97.55% in the prediction of anxiety level in the test results. The accuracy, sensitivity, and specificity of the LR classifier were reported as 78.07%, 14.87%, and 96.81%, respectively. However, the k-nearest neighbor algorithm showed the same performance as the RF, with accuracy, sensitivity, and specificity of 91.30%, 73.55%, and 96.57%, respectively. Naive Bayes showed an accuracy of 73.53%, a sensitivity of 24.13%, and a specificity of 94.11% when predicting the anxiety state of the test respondents.
Among these six classifiers, the best result was achieved by the SVM (linear) algorithm, showing that the accuracy was 91.49%, the sensitivity was 67.77%, and the specificity was 98.53%. According to Cohen's kappa value of SVM classifier, the data showed that the reliability exceeded 60% (k = 0.7333).
This violin plot shows the relationship of seven classifiers to accuracy. The shaded areas detail the distribution of the data in each classifier. In terms of depression, [Figure 2]a shows that RF provided the highest mean accuracy, followed by KNN, LR, LDA, SVM, and NB. [Figure 2]b shows that SVM provided the highest mean accuracy for anxiety, followed by RF, KNN, LDA, LR, and NB. Unlike the boxplot, the entire distribution of the 10-fold accuracy can be visualized in this violin plot [Figure 2].
|Figure 2: Violin plot showing depression accuracy (a) and anxiety accuracy (b) in each machine learning classifiers|
Click here to view
| Discussion|| |
To the best of our knowledge, this is the first study to apply several machine classifiers to predict the level of depression and anxiety status among university students during the first wave of COVID-19 in Bangladesh.
In Bangladesh, mental health illnesses were not as important as all other public health issues such as malnutrition, and mental health issues are now becoming a major issue around the world. Few studies have been conducted on mental disorders among college students., Thus, the main focus of the study was to predict the state of mental disorders such as depression and anxiety among Bangladeshi university students. To fulfill our study objective, this study used six well-known ML algorithms. All models were trained based on 10-fold cross-validation on the training set, and performance was estimated in the testing set.
The prevalence of severe depression and severe anxiety was higher for female than for male. A recent systematic review analysis depicts that female gender was significantly important to increase the depression symptom. This study found a high prevalence of depression and anxiety problem for teenage students who were currently in the study. There are many psychological reasons for higher percentage of severe mental health problem among students. For instances, students felt insecure about their future career during this pandemic situation. Lower academic performance is also an important and significant indicator for increasing the mental health problem among university students.
From supervised model comparison, the best accuracy was achieved by the RF algorithm for depressed sectors based on various performance parameters. But for anxiety states, the SVM showed the best predictive estimate with 91% accuracy. Compared to previous research SVMs, it performed with a reasonable level of accuracy among all classifiers. In other areas of public health, such as malnutrition and anemia in children, the RF has great predictive potential., In research related to mental health or medical sector, RF and support vector algorithm were used mostly to predict psychiatric disorder and disease outcome, respectively, by several researchers.,,,, Finally, this study proposed that the classification of RFs and vector machines be extended where Bangladesh's coerced concern is predicting mental health problems, for example, depression and anxiety.
Almost all studies will have few impediments, and current research is not without limitations that must be taken seriously during data interpretation. First, the nature of this research is cross-sectional, for which it is quite impossible to provide a causal relationship. During the lockdown, data collection is really impossible, for that reason, we have conducted the online structure questionnaire method. Second, the study was conducted by interviewing online with a small number of students from four universities, not representing university students from the whole country. Finally, due to the online survey, few students were unable to complete the questionnaire by their own hands and provide accurate information due to mental shame.
However, this investigation established that ML algorithms can be used to predict psychological state malady supported common risk factors, which can assist within the development of interventions to prevent severe psychological state issues among students, particularly university students in Bangladesh.
| Conclusions|| |
This study focuses mainly on the comparison of the performance of ML-based algorithms to predict depression and anxiety among university students during the first wave of the COVID-19 pandemic in Bangladesh. This research is also important for policymakers since Bangladesh has the second-highest temporary university closed in the world in an attempt to contain the spread of the COVID-19 pandemic. During the time of the first wave of the COVID-19 pandemic, students faced severe mental health problems. Based on the study findings, the prevalence of severe depression and severe anxiety among university students was 26.7% and 22%, respectively. In summary, we conclude that among ML classifiers, a RF is best for predicting depression, and an SVM is best for predicting anxiety. Finally, we suggest that RF and SVMs are the core interests of researchers in predicting depression and anxiety among university students, respectively.
The authors would like to express the deepest gratitude to all the enthusiastic and volunteer respondents who participated in this research. Many thanks also to the developers who developed 9-item Patient Health Questionnaire depression scale (PHQ-9) and 7-item GAD-7 anxiety scale.
Declaration of respondent consent
The authors certify that they have obtained all appropriate university student consent.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Eurosurveillance Editorial Team. Note from the editors: World Health Organization declares novel coronavirus (2019-nCoV) sixth public health emergency of international concern. Euro Surveill 2020;25:1-2. 10.2807/1560-7917.ES.2020.25.5.200131e.
Bhattacharyya R, Chatterjee S, Bhattacharyya S, Gupta S, Das S, Banerjee B. Attitude, practice, behavior, and mental health impact of COVID-19 on doctors. Indian J Psychiatry 2020;62:257. [Full text]
Atchison CJ, Bowman L, Vrinten C, Redd R, Pristera P, Eaton JW, et al.
Early Perceptions and Behavioural Responses during the COVID-19 pandemic: a cross-sectional Survey of UK Adults. BMJ Open 2021;11:e043577.
Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C, Imai N, et al.
Estimates of the severity of coronavirus disease 2019: A model-based analysis. Lancet Infect Dis 2020;20:669-77.
Rajabimajd N, Alimoradi Z, Griffiths M. Impact of COVID-19-related fear and anxiety on job attributes: A systematic review. Asian J Soc Health Behav 2021;4:51. [Full text]
Pramukti I, Strong C, Sitthimongkol Y, Setiawan A, Pandin MG, Yen CF, et al.
Anxiety and suicidal thoughts during the COVID-19 pandemic: Cross-country comparative study among Indonesian, Taiwanese, and Thai University Students. J Med Internet Res 2020;22:e24487.
Sumathi B. Prediction of mental health problems among children using machine learning techniques. Int J Adv Comput Sci Appl 2016;7:1-6.
Abdallah AR, Gabr HM. Depression, anxiety and stress among first year medical students in an Egyptian Public University. Int Res J Med Med Sci 2014;2:11-9.
Alim SA, Rabbani MG, Karim E, Mullick MS, Mamun AA, Fariduzzaman. Assessment of depression, anxiety and stress among first year MBBS students of a public medical college, Bangladesh. Bangladesh J Psychiatry 2017;29:23-9.
Beiter R, Nash R, McCrady M, Rhoades D, Linscomb M, Clarahan M, et al.
The prevalence and correlates of depression, anxiety, and stress in a sample of college students. J Affect Disord 2015;173:90-6.
Alimoradi Z, Ohayon MM, Griffiths MD, Lin CY, Pakpour AH. Fear of COVID-19 and its association with mental health-related factors: Systematic review and meta-analysis. BJPsych Open 2022;8:e73.
Gewin V. Five tips for moving teaching online as COVID-19 takes hold. Nature 2020;580:295-6.
Sahu P. Closure of universities due to coronavirus disease 2019 (COVID-19): Impact on education and mental health of students and academic staff. Cureus 2020;12:e7541.
Boyraz G, Legros DN. Coronavirus disease (COVID-19) and traumatic stress: Probable risk factors and correlates of posttraumatic stress disorder. J Loss Trauma 2020;25:503-22.
Islam MS, Ferdous MZ, Potenza MN. Panic and generalized anxiety during the COVID-19 pandemic among Bangladeshi people: An online pilot survey early in the outbreak. J Affect Disord 2020;276:30-7.
Islam MS, Sujan MS, Tasnim R, Sikder MT, Potenza MN, van Os J. Psychological responses during the COVID-19 outbreak among university students in Bangladesh. PLoS One 2020;15:e0245083.
Rubin GJ, Wessely S. The psychological effects of quarantining a city. BMJ 2020;368:m313.
Brooks SK, Webster RK, Smith LE, Woodland L, Wessely S, Greenberg N, et al.
The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. Lancet 2020;395:912-20.
Zandifar A, Badrfam R. Iranian mental health during the COVID-19 epidemic. Asian J Psychiatr 2020;51:101990.
Mamun MA, Hossain MD, Griffiths MD. Mental health problems and associated predictors among Bangladeshi students. Int J Ment Health Addict 2022;20:657-71.
Alghamdi M, Al-Mallah M, Keteyian S, Brawner C, Ehrman J, Sakr S. Predicting diabetes mellitus using SMOTE and ensemble machine learning approach: The Henry Ford ExercIse Testing (FIT) project. PLoS One 2017;12:e0179805.
Kotsiantis SB, Zaharakis ID, Pintelas PE. Machine learning: A review of classification and combining techniques. Artif Intell Rev 2006;26:159-90.
Priya A, Garg S, Tigga NP. Predicting anxiety, depression and stress in modern life using machine learning algorithms. Procedia Comput Sci 2020;167:1258-67.
Sau A, Bhakta I. Screening of anxiety and depression among the seafarers using machine learning technology. Inform Med Unlocked 2019;16:100149.
Kroenke K, Spitzer RL, Williams JB. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med 2001;16:606-13.
Kroenke K, Spitzer RL. The PHQ-9: A new depression diagnostic and severity measure. Psychiatr Ann 2002;32:509-15.
Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Intern Med 2006;166:1092-7.
Haque M, Das C, Ara R, Alam M, Ullah S, Hossain Z. Prevalence of generalized anxiety disorder and its effect on daily living in the rural community of Rajshahi. J Teach Assoc 2018;27:14-23.
Zhang T, Su J, Xu Z, Luo Y, Li J. Sentinel-2 satellite imagery for urban land cover classification by optimized random forest classifier. Appl Sci 2021;11:543.
Junli C, Licheng J. Classification Mechanism of Support Vector Machines. WCC 2000-ICSP 2000 2000 5th
International Conference on Signal Processing Proceedings 16th
World Computer Congress 2000; 2000.
Güvenç E, CetinGÇ, Koçak H. Comparison of KNN and DNN classifiers performance in predicting mobile phone price ranges. Adv Artif Intell Res 2021;1:19-28.
Vembandasamy K, Sasipriya R, Deepa E. Heart diseases detection using naive bayes algorithm. Int J Innov Sci Eng Technol 2015;2:1-4.
Bayram N, Bilgel N. The prevalence and socio-demographic correlations of depression, anxiety and stress among a group of university students. Soc Psychiatry Psychiatr Epidemiol 2008;43:667-72.
Kumari R, Langer B, Jandial S, Gupta RK, Raina SK, Singh P. Psycho-social health problems: Prevalence and associated factors among students of professional colleges in Jammu. Indian J Community Health 2019;31:43-9.
Lun KW, Chan CK, Ip PK, Ma SY, Tsai WW, Wong CS, et al.
Depression and anxiety among university students in Hong Kong. Hong Kong Med J 2018;24:466-72.
Akanni O, Olashore A, Fela-Thomas A, Khutsafalo K. The psychological impact of COVID-19 on health-care workers in African countries: A systematic review. Asian J Soc Health Behav 2021;4:85. [Full text]
Agnafors S, Barmark M, Sydsjö G. Mental health and academic performance: A study on selection and causation effects from childhood to early adulthood. Soc Psychiatry Psychiatr Epidemiol 2021;56:857-66.
Srividya M, Mohanavalli S, Bhalaji N. Behavioral modeling for mental health using machine learning algorithms. J Med Syst 2018;42:88.
Talukder A, Ahammed B. Machine learning algorithms for predicting malnutrition among under-five children in Bangladesh. Nutrition 2020;78:110861.
Khare S, Kavyashree S, Gupta D, Jyotishi A. Investigation of nutritional status of children based on machine learning techniques using Indian demographic and health survey data. Procedia Comput Sci 2017;115:338-49.
Aguiar-Pulido V, Gestal M, Fernandez-Lozano C, Rivero D, Munteanu CR. Applied computational techniques on schizophrenia using genetic mutations. Curr Top Med Chem 2013;13:675-84.
Aguiar-Pulido V, Seoane JA, Rabuñal JR, Dorado J, Pazos A, Munteanu CR. Machine learning techniques for single nucleotide polymorphism-disease classification models in schizophrenia. Molecules 2010;15:4875-89.
Vivian-Griffiths T, Baker E, Schmidt KM, Bracher-Smith M, Walters J, Artemiou A, et al.
Predictive modeling of schizophrenia from genomic data: Comparison of polygenic risk score with kernel support vector machines approach. Am J Med Genet B Neuropsychiatr Genet 2019;180:80-5.
Li C, Yang C, Gelernter J, Zhao H. Improving genetic risk prediction by leveraging pleiotropy. Hum Genet 2014;133:639-50.
Acikel C, Aydin Son Y, Celik C, Gul H. Evaluation of potential novel variations and their interactions related to bipolar disorders: Analysis of genome-wide association study data. Neuropsychiatr Dis Treat 2016;12:2997-3004.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]