PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

CHAPTER 1

INTRODUCTION
Mental health disorder, disrupts person’s thinking, feeling, mood, ability and daily
functioning. Nowadays many people or professionals are suffering from mental disorder.

In the modernized world, working professionals are under lots of pressure for reasons
like peer pressure, short deadlines, competition. All these things contribute to building up
mental stress, which slowly leads to mental health disorders.

Globally, mental disorders are leading causes of disability. Mental health problems in
the workplace have serious effects not only for the individual but also for the productivity and
competitiveness of businesses and thus the economy and society as a whole. Employees’
mental health status affects employees’ performance and rates of illness, absenteeism and staff
turnover and, It affects work performance in terms of increase in error rates, poor decision-
making, loss of motivation and commitment, tension and conflicts between colleagues.

Mental health prediction is conductive to detecting mental disorders in advance,

reducing the incidence of serious mental illnesses, and facilitating the health system to provide
people with targeted health care services. Mental illness has been and remains a major cause
of disability, malfunction, and even violence and crime. Tracking mental health on a large
scale is a major challenge in terms of resources. Traditional methods of monitoring mental
health are expensive, intrusive, and often geared toward serious mental disorders. More
importantly, these methods do not scale to a large population of varying demographics, and
are not particularly designed for those in the early stages of developing mental health problems.
In the meantime, effective intervention for mental health has been severely limited by the
availability of resources in healthcare professionals and treatments.

Mental health disorder is usually diagnosed based on the individual self-report that
requires questionnaires designed for the detection of the specific patterns of feeling or social
interactions. With proper care and treatment, many individuals will hopefully be able to
recover from mental health disorder or emotional disorder. Our motive is to raise awareness
about mental health disorders. And early prediction of mental health disorder.
Dept. of ISE, CIT, Gubbi. Page 1
PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

1.1 Problem Statement

The mental health includes our emotional, psychological and social well-being. mental
disorders are leading causes of disability and It affects a thinking ability of an employee. Mental
health problem that undoubtedly impacts emotions, reasoning, and social interaction of a person.
These issues have shown that mental illness gives serious consequences across societies and
demands. Early prediction of the mental health disorder is difficult.

1.2 Objectives

To Raise awareness about the mental health continuum.

promote help seeking behaviors and emotional well-being practices.
And prevent suicide through individual education and outreach events.
Promoting active participation and encouraging personal development

1.3 Methodology

There are 4 stages to predict Mental health disorder.

Planning stage, In this stage we have to plan about prediction of mental health problem.
The queries such as Machine Learning Algorithms in Mental Health, Psychiatric Medical with
Machine Learning Techniques, and Machine Learning in Predicting Mental Health Problems
have been analyzed and planned in this stage.

The analysis phase is started by finding out and investigating the performance of the
machine learning approaches that were used to diagnose or predict mental health problems.

The discussion phase will begin by reviewing the machine learning algorithms used by
the researchers in their experiments to predict the mental problems. Mental health problems
will be divided and categorized into several parts. Then, the performance for the machine
learning techniques will be described and further analyzed in this phase.

The conclusions related to the topic will be highlighted based on the findings and
discussion. Moreover, the prediction of the mental health problems by using machine learning
approaches will be generalized and summarized.

Dept. of ISE, CIT, Gubbi. Page 2

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

Chapter 2

TYPES OF MENTAL HEALTH PROBLEMS

Mental illness can affect the cognition, emotion, and behaviour among the people. For
children, their ability to learn could be interfered by mental disorders. Besides that, mental
illness can cause inconvenience to the adults, especially in their families, workplaces, and in
the society. There are many types of mental disorders commonly known as mood disorder,
schizophrenia, depression, bipolar disorder, and anxiety.

2.1 The mood disorder

The mood disorder is serious changes in mood in the form of emotional inconsistency
or abrupt changes/amplification of certain specific emotions, which can be feeling extremely
sad or feeling irritable. When they start interfering or disrupting normal life activities, we call
it mood disorder. And mood disorder can be further characterized in the forms of depression
disorder, mania, hypomania and bipolar disorder. Depression disorder as the name suggests is
synonym to depression, negative feelings start influencing the daily life of a person. Which
includes mood swings, losing interest in daily life events, feeling apathy, hard time sleeping,
losing appetite or feeling overly hungry. Depression in later phases may lead to suicidal
tendencies. Mania is hyperactive state of a person where a person has an excess of both
physical and mental energy, whose symptoms are feeling restless and failing to sit still, easily
getting distracted, not able to focus on a particular thing, having a hard time sleeping, talking
too much.

2.2 Anxiety disorder

Anxiety is an emotional response to a future event, which is usually in the form of fear.
Anxiety is a normal emotion but an anxiety disorder is an entirely different case. In which we
feel an excessive amount of fear and anxiety for no reason whatsoever, excess anxiety can
make people skip meetings, avoid social interactions and much more. We can be further
categorized as General anxiety disorder, phobia anxiety disorder, and social anxiety disorder.

Dept. of ISE, CIT, Gubbi. Page 3

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

2.3 Schizophrenia

Schizophrenia is a mental illness that interrupted by events of psychotic symptoms,

which are hallucinations and delusions. Hallucinations are experiences that are not
comprehensible to others. Meanwhile, delusions are impressions that are held by the patients
although contradicted by the rational and real arguments. Schizophrenia is often diagnosed by
symptoms such as social withdrawal, irritability, and increasing strange behaviours. Studies of
whether an early diagnosis of such symptoms and intervention could improve the outcomes
are still in progress.

2.4 Depression

The primary symptom of depression is an interference of the mood, which is usually

severe sadness. Sometimes, anger, irritability, and loss of interests might dominate the
symptoms of the depression. In terms of physiological symptoms, sleep disturbance, appetite
disturbance, and decreased in energy are commonly shown across cultures. The cognitive
symptoms such as slow thinking, suicidal thoughts, and guilt might occur among the patients.
Most of the individuals that suffer from depression will have recurrence episodes. Many
individuals do not recover completely and they might have a form of chronic mild depression.

2.5 Bipolar disorder

Bipolar disorder is another mental disorder identified by the episode of mania and
depression. Sometimes, there is an episode mixed with both mania and depression. Mania is
known by irritability, increased in energy, and decreased need for sleep. Individuals that
experience mania often exhibit reckless behaviours. Meanwhile, a depressive episode for
bipolar disorder is almost the same as the depression symptoms. Some studies report some
recovery to baseline functioning between episodes; however, many patients will have residual
symptoms that cause impairment.

Dept. of ISE, CIT, Gubbi. Page 4

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

2.6 Addiction/Substance Abuse

Addiction is a complex brain disease manifested by compulsive substance use despite

harmful consequences. People with addiction have an intense focus on using an item such as
alcohol or drugs, to the point that it takes over their lives. They often forsake school, jobs and
personal relationships to fuel their habit.

Dept. of ISE, CIT, Gubbi. Page 5

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

CHAPTER 3

MACHINE LEARNING TECHNIQUES USED

Machine learning is a subset of AI, which allows the system to learn from previous
actions and improve without giving explicit commands. Machine learning can be categorized
into two types based on the type of learning or the data used.
In Supervised learning, we have labelled data where labelled means both input and
output are known to us. Whereas in Unsupervised learning we have unlabeled data, which
means we are given values of independent variables and value of dependent variables are
absent. We will only use supervised learning algorithms as the dataset used is labelled.
3.1 Logistic regression
Logistic regression is one of the most popular Machine Learning algorithms, which
comes under the Supervised Learning technique. It is used for predicting the categorical
dependent variable using a given set of independent variables.
It uses the sigmoid function when predicting the dependent variable using one or more
dependent variables. The dependent variable usually is a binary number that ranges
between 0 and 1.
Logistic Regression can be used to classify the observations using different types of
data and can easily determine the most effective variables used for the classification.
3.2 Random Forest
Random Forest is a classifier that contains a number of decision trees on various subsets
of the given dataset and takes the average to improve the predictive accuracy of that
dataset.
It is an ensemble model which means it uses many machine learning algorithms to
increase its performance as compared to other machine learning algorithms. Random
forest randomly picks up a subset from the training data set and generate various
decision trees. It will predict the class of test class objects using the decision trees.
It takes less training time as compared to other algorithms.

Dept. of ISE, CIT, Gubbi. Page 6

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

3.3 Decision tree

A decision tree is a supervised machine-learning algorithm so that means the data set
should be labelled. In Decision tree algorithm, the classification is done based on a set
of rules.
In a decision tree, a node will represent a feature, the branch will represent a rule and
the leaf node will represent the outcome. It can be represented in a tree-like structure
which provides higher stability and accuracy.
In Decision tree Algorithm following steps are taken in the first step a tree will be
constructed which will have input features as its nodes.
In the next step, it will select a feature from the input features for predicting the output,
which gives the highest information gain. Now use the above steps for the creation of
subtrees by making use of the features, which are not used earlier.
They used decision tree to find features, which contribute to the mental health disorders
and the degree to which they contribute to the mental health disorder, by using feature
importance.

3.4 Support vector machine

Support Vector Machine or SVM is one of the most popular Supervised Learning
algorithms, which is used for Classification as well as Regression problems. However,
primarily, it is used for Classification problems in Machine Learning.
SVM takes data points as input and gives the output as a hyperplane. It divides the
classes using a plane (hyperplane) also known as the decision boundary. Where the
decision boundary must maximize the distance of the nearest element of each class.
Decision boundary separates the points into different classes.

3.5 Naive bayes

Naïve Bayes Classifier is one of the simple and most effective Classification algorithms
which helps in building the fast machine learning models that can make quick
predictions.
It is a probabilistic classifier, which means it predicts on the basis of the probability of
an object.

Dept. of ISE, CIT, Gubbi. Page 7

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

It is a probability-based classifier which applies Bayes theorem. For Naïve Bayes

features should have strong independence.

3.6 K-Nearest Neighbour

K-NN algorithm stores all the available data and classifies a new data point based on
the similarity. This means when new data appears then it can be easily classified into a
well suite category by using K- NN algorithm.
K-NN algorithm can be used for Regression as well as for Classification but mostly it
is used for the Classification problems.
KNN algorithm at the training phase just stores the dataset and when it gets new data,
then it classifies that data into a category that is much similar to the new data.
As it is a supervised learning algorithm, we will need labelled data for training the
model. When classifying a data point, we look at K nearest neighbours of the data the
majority will decide which class the data point will belong.

Dept. of ISE, CIT, Gubbi. Page 8

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

CHAPTER 4
LITERATURE REVIEW
World Health Organization (WHO) reports the region-wise status of different barriers
in diagnosing mental health problems and encourages doctors to be equipped with the scientific
knowledge to address the issue of mental health. Now, there are various techniques to predict
the state of mental health due to advancement of technology. Research in the field of mental
health has increased recently and contributed to the information and publications about
different features of mental health, which can be applied in a wide range of problems. Many
steps are involved in diagnosing mental health problems. Generally, the diagnosis will begin
with a specific interview that is filled with questions about symptoms, medical history, and
physical examination. Besides that, psychological tests and assessment tools are also available
and are used to diagnose a person for mental health problems. There are several types of
research carried out to investigate and examine the movements of the face to identify certain
mental disorders.
Machine learning is an application of artificial intelligence (AI), which implements
systems with the capability to learn and improve from experience without being explicitly
programmed. Machine learning has offered essential advantages to a wide range of areas. In
the medical field, machine learning algorithms have been used to discover new drugs, perform
radiology analysis, predict epidemic outbreaks, and diagnose diseases. Machine learning
approaches continue to develop rapidly. Powerful algorithms and more advanced neural
networks, decision trees, and others were introduced and applied to solve more complicated
medical diagnosis problems.
Generally, machine learning algorithms are tools to analyze the massive medical
datasets. They are utilized as tools in assisting for medical diagnosis as they became more
reliable in their performance.

Dept. of ISE, CIT, Gubbi. Page 9

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

Train data

Data processing/Feature
selection

Train ML algorithm

Model building

Models Test data

Model Evaluation

Finding Feature
Importance

Figure 1: Process of the systematic literature review for this topic.

The dataset contains more than 70 attributes, which contain both personal and professional
features of the employees. In the feature selection part of data processing, we select seven
attributes out of 70 attributes, which contribute most to the mental disorder or can be used to
predict the mental health disorder. The feature selection is done by carefully scrutinizing done
on the base of past works/papers and some research on disorders. After the feature selection
comes the classification part in which different ML algorithms are applied to the selected
features to predict the mental disorders and at last we apply feature importance to find the
attributes which are most useful in predicting the mental health disorders.

Dept. of ISE, CIT, Gubbi. Page 10

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

CHAPTER 5
IMPLIMENTATIONS
The experiments of the classification are conducted with various machine learning
models. It is undeniable that machine learning models such as random forest and support vector
machine have been the most popular choice to be applied in the experiments. This is because
random forest and support vector machine at most of the time are able to provide an excellent
performance in terms of the accuracy.
5.1 Small Sample Size
Machine learning can exhibit robustness when analyzing the large sample size, certain
approaches can perform with a small sample without compromising the accuracy depending
on the settings toward the model applied in the experiments. small sample is common in the
field of mental health because of the cost that is related to the data collection that involves the
human participants and the experimental rules with different conditions are still under
development.

5.2 Limited Exploration in Deep Learning

It is believed that deep learning algorithms have been successful for a few
applications, especially in healthcare domains. However, there is still limited exploration in
the use of deep learning algorithms for mental health.

5.3 Lack of Real-Life Testing

Although machine learning can show the researchers about the prediction on mental
health, there is still a lack of testing being applied in real life due to several reasons. Many
medical professionals still doubt the accuracy of automated methods such as machine learning,
as well as issues of consistency and difficulty when applying the machine learning predictive
systems to real-world medical practices.

Dept. of ISE, CIT, Gubbi. Page 11

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

5.4 Exploration in Deep Learning

The success of applying machine learning approaches in mental health prediction can
be expanded to include deep learning approaches. Such approaches could even predict mental
health problems together with diagnosis of other chronic diseases such as cancer, diabetes, and
others.

5.5 High-Quality Data

In order to develop more accurate predictive tools, data such as sociodemographics,

speech, medical report profiles, and facial expressions of the patients can be recorded or taken
via photography combined with magnetic resonance imaging of the brain. In this approach, the
data have a larger volume where the deep learning algorithms can be useful and applied.
Obtaining such a detailed and large data set shows a challenge for the mental health field and
requires immediate collaboration among institutes and organizations.

5.6 Accurate Predictive Tools

The application of new models to predict the clinical results should be given the
research opportunity. Besides, web-based predictors and medical analytics tools should be
developed to transform the effective predictive models into useful clinical decision systems
such as for identifying the different types of mental disorders, medication plans, as well as
preventive plans.

5.7 Explainable Model

Performance of machine learning models and being explainable are necessary for
mental health problems. Medical professionals need to understand the underlying system of
prediction and classification very well before practising it in the real world and with patients.
Making the results obtained by these models understandable should be the main priority toward
establishing reliable systems.

Dept. of ISE, CIT, Gubbi. Page 12

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

RESULTS
After applying various machine-learning algorithms when predicting mental health
disorder with selected attributes which are
1. Whether the company is a Tech company or not.
2. Age of the employee.
3. Gender of the employee.
4. Family history of mental health disorders if any.
5. Personal history of a mental health disorder.
6. Mental health benefits or care provided by the employer.
7. Discussing mental health status with the employer.
We got the following results.
positive outputs
S.no Algorithms Accuracy (%) Precision Recall correctly predicted
by the classifier
1. KNN 74 76 82 82
2. SVM 76 75 88 88
3. Logistic Regression 84 82 94 94
4. Decision Tree 84 83 92 92
5. Random Forest 77 81 80 80
6. Naïve bayes 79 78 90 90

TABLE 1. COMPARISON OF CLASSIFICATION MODELS

As shown in TABLE 1 logistic regression and decision tree shows the highest accuracy when
predicting mental health disorders in employees of both tech and non-tech companies.

Dept. of ISE, CIT, Gubbi. Page 13

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

CONCLUSION
Many different techniques and algorithms had been introduced and proposed to test
and solve the mental health problems.
After analyzing, we found that decision tree classifier has the best performance. As it
has the best accuracy and precision with accuracy 84% and precision 83 followed by logistic
regression with 84% accuracy and 82 precision followed by Naïve Bayes with 79% accuracy
and 78 precision, random forest with 77% accuracy and 81 precision, SVM with 76% accuracy
and 75 precision and KNN has the worst performance with 74% accuracy and 76 precision.
Also, Feature importance of the selected features showed that a history of mental health
disorder contributes most during disorder prediction followed by family history. It was also
found that rest of the features contributes bare minimum to the prediction with gender as their
top rest of the features which includes mental health benefits or care provided by the employer,
age and discussing mental health status with the employer barely makes any contribution to
the prediction of mental health disorder.
It may also help distinguish and classify the mental health problems among patients for
further treatment.

Dept. of ISE, CIT, Gubbi. Page 14

PREDICTING MENTAL HEALTH DISORDER USNG MACHINE LEARNING 2021-22

REFERENCES
1. U. S. Reddy, A. V. Thota, and A. Dharun.” Machine Learning Techniques for Stress
Prediction in working employees.”
2. Jetli Chung and Jason Teo “Mental Health Prediction Using Machine Learning:
Taxonomy, Applications, and Challenges”
3. Praveen Kumar “Mental Health Prediction Based on Social Behaviour”
4. Rahul Katarya “Predicting Mental health disorders using Machine Learning for
employees in technical and non-technical companies”
5. https://www.javatpoint.com/logistic-regression-in-machine-learning
6. https://www.javatpoint.com/k-nearest-neighbor-algorithm-for-machine-learning
7. https://www.javatpoint.com/machine-learning-naive-bayes-classifier
8. https://www.javatpoint.com/machine-learning-random-forest-algorithm
9. https://www.javatpoint.com/machine-learning-decision-tree-classification-
algorithm

Dept. of ISE, CIT, Gubbi. Page 15