KWARA STATE UNIVERSITY, MALETE

The University for Community Development

Faculty of Information and Communication Technology

DEVELOPMENT OF A MACHINE LEARNING MODEL TO PERSONALIZE

VACCINATION SCHEDULES BASED ON INDIVIDUAL HEALTH FACTORS.
BY

ADEJESU OREOLUWA EMMANUEL

August 2024
 DEVELOPMENT OF A MACHINE LEARNING MODEL TO PERSONALIZE
VACCINATION SCHEDULES BASED ON INDIVIDUAL HEALTH FACTORS.

BY

ADEJESU OREOLUWA EMMANUEL

A RESEARCH PROJECT SUBMITTED TO THE DEPARTMENT OF

COMPUTER SCIENCE, FACULTY OF INFORMATION AND
COMMUNICATION TECHNOLOGY, KWARA STATE UNIVERSITY,
MALETE, IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
AWARD OF BACHELOR OF SCIENCE (B.Sc.) DEGREE IN COMPUTER
SCIENCE.

August 2024

ii
 DECLARATION
I hereby declare that this research work titled “Development of a machine learning
model to personalize vaccination schedules based on individual health factors” is my
own work and has not been submitted by any other person for any degree or qualification
at any higher institution. I also declare that the information provided therein are mine and
those that are not mine are properly acknowledged.

__________________________

________________________
Names Adejesu Oreoluwa Emmanuel
Name of student Signature and Date

iii
 CERTIFICATION

This is to certify that the research project titled Development of a machine learning
model to personalize vaccination schedules based on individual health factors was
carried out by “Adejesu Oreoluwa Emmanuel”. The project has been read and
approved as meeting the requirements for the award of Bachelor of Science (B.Sc.)
Degree in Computer Science in the Department of Computer Science, Faculty of
Information and Communication Technology, Kwara State University, Malete.

______________________ ___________________
Dr. R.M. Isiaka Signature/Date
Supervisor

_______________________ ____________________
Dr. (Mrs.) R.S. Babatunde Signature/Date
Head of Department

_______________________ _____________________
External Examiner Signature/Date

iv
 DEDICATION
This Project is dedicated to Almighty GOD, the beginning and the end who has been with
me since my birth till the moment. Also, to my parents (please put your family people),
my guardians, supervisor and my boss at (People you can call boss) for their supports,
guidance and prayers.

v
 ACKNOWLEDGEMENT
All praise and adoration belong to Almighty GOD for his mercy and protection over me
throughout my program in the university.
I acknowledge the efforts of my parents Mr/Mrs C.O. ADEJESU, may GOD Continue to
bless them . My sincere appreciation also goes to my loving and caring brothers and
sisters starting from (people you love) for their roles and Muhammed Jamiu Olanrewaju
for his courageous words towards the success of this program, and thanks to entire family
and its community in general. May Allah reward them all abundantly, Furthermore, I
acknowledge the support of my friends from (friends). May Almighty God be with them
and crown their efforts with success.
I appreciate my colleagues in the university, (class friends) and my entire class mates.
May He answer our prayers and crown all our efforts with success. The school authority
is also inclusive, for creating an opportunity and avenue for us to be exposed to the
outside world.
My profound gratitude goes to my supervisor, Dr. R.M. Isiaka, who did all he could to
make this report a successful one. My appreciation also goes to all lecturers in the
department.

vi
 Table of content
CERTIFICATION..............................................................................................................iv

DEDICATION.....................................................................................................................v

ACKNOWLEDGEMENT..................................................................................................vi

CHAPTER ONE..................................................................................................................1

INTRODUCTION...............................................................................................................1

1.1 Background of the study.........................................................................................1

1.2 Statement of the problem........................................................................................2

1.3 Aim and objectives.................................................................................................3

1.4 Scope of study.........................................................................................................3

1.5 Significance/justification of the study....................................................................3

1.6 Definition of terms..................................................................................................4

CHAPTER TWO.................................................................................................................7

LITERATURE REVIEW....................................................................................................7

2.1 Review of related terms..........................................................................................7

2.2 Related Work..........................................................................................................9

CHAPTER THREE...........................................................................................................16

SYSTEM DESIGN AND ANALYSIS..............................................................................16

3.1 Data Acquisition and Planning.............................................................................16

Dataset.........................................................................................................................16

Dataset preprocessing and integration........................................................................17

Feature Selection.........................................................................................................18

3.2 Classification algorithm development..................................................................18

vii
 3.3 User interface development:.................................................................................22

CHAPTER FOUR..............................................................................................................29

RESULTS AND DISCUSSIONS......................................................................................29

Minimum requirements...............................................................................................29

Best System Requirements.........................................................................................29

Interfaces.....................................................................................................................30

Functionality...............................................................................................................31

4.2 Discussion.............................................................................................................34

CHAPTER FIVE...............................................................................................................35

SUMMARY, CONCLUSION, AND RECOMMENDATION.........................................35

Summary.....................................................................................................................35

Conclusion..................................................................................................................35

Recommendations.......................................................................................................36

References...................................................................................................................38

Appendix:....................................................................................................................41

viii
LIST OF FIGURES
Vaccination dataset from WHO…………………………………………………….17
Feature selection and data selection………………………………………………...18
Random forest regressor training………………………………………………….. 19
Tkinter user interface design………………………………………………………..25
Algorithm integration……………………………………………………………….
26
Application homepage……………………………………………………………... 30
Interaction when the ward is predicted to have taken the vaccine………………….32
Country dropdown menu…………………………………………………………... 33

ix
LIST OF APPENDICES
Appendix 1 …………………………………………………………………………41

x
 ABSTRACT
This project aims to develop a personalized vaccination scheduling system using machine
learning classification algorithms, leveraging data from the World Health Organization
(WHO). Three datasets detailing vaccination schedules for rubella, mumps, and measles
were collected and integrated, encompassing vaccines recommended for various target
populations, including adults, health workers, pregnant women, and risk groups. The
project emphasizes the importance of data preprocessing, feature selection, and the
application of classification algorithms such as Decision Trees and Support Vector
Machines (SVM) for accurate predictions. A user-friendly interface is developed using
Python to facilitate interaction with the classification model, enabling users to input data
and receive vaccination schedule predictions effectively.

xi
 CHAPTER ONE

INTRODUCTION

1.1 Background of the study

The concept of personalized vaccination schedules arises from the need to enhance the
effectiveness and safety of immunizations by tailoring them to individual health factors.
Traditional vaccination schedules are typically based on population-wide data and
guidelines, which do not account for the variability in individual health status, genetic
predispositions, and other personal factors. Recent advancements in medical technology
and data analytics have paved the way for more personalized approaches, potentially
revolutionizing public health and individualized patient care (Topol, 2019).
The development of a machine to personalize vaccination schedules heavily relies on the
integration of comprehensive health data and machine learning algorithms. This approach
extends beyond traditional "one‐size‐fits‐all" treatments, offering the potential to halt the
autoimmune process, preserve beta‐cell function, and optimize therapeutic strategies
during the critical honeymoon period. (Kannan et al., 2023). For instance, neural
networks and other AI models can be trained to optimize vaccination timing and dosage
by continuously learning from new data (Chen et al., 2020). This approach ensures that
each vaccination plan is uniquely tailored to maximize efficacy and minimize adverse
reactions.
Building the technological infrastructure for personalized vaccination scheduling
involves several critical components. First, a robust data collection system is necessary to
gather and integrate data from various sources, such as genomic databases, wearable
health devices, and patient health records. Second, advanced machine learning models
must be developed and validated to ensure their predictive accuracy and reliability. These
models require continuous updates and training with new data to maintain their

1
effectiveness. Additionally, implementing such a system in clinical settings necessitates
collaboration between healthcare providers, data scientists, and IT specialists to ensure
seamless integration and usability (Lee & Yoon, 2021).
While the benefits of personalized vaccination schedules are significant, ethical
considerations and data privacy issues must be carefully addressed. The use of personal
health data requires stringent measures to protect patient privacy and secure sensitive
information. Ensuring informed consent and transparency in how data is used and shared
is crucial to maintaining patient trust. Furthermore, there is a need to address potential
biases in data collection and algorithm design to avoid disparities in healthcare outcomes.
Ethical frameworks and regulatory guidelines should be established to oversee the
development and deployment of these technologies (Morley et al., 2020).
The future of personalized vaccination schedules holds great promise for improving
public health outcomes. As technology advances, the accuracy and feasibility of these
systems will continue to improve, potentially reducing the incidence of vaccine-
preventable diseases and enhancing herd immunity. Personalized vaccination schedules
could also lead to more efficient use of healthcare resources by minimizing adverse
reactions and optimizing immunization programs. Moreover, the insights gained from
personalized vaccination data could inform broader public health strategies and vaccine
development, ultimately contributing to a more resilient and responsive healthcare
system (Gordon et al., 2021).

1.2 Statement of the problem

One significant problem in public health is vaccine hesitancy, which is partly driven by
concerns over adverse reactions and a lack of trust in standardized vaccination schedules.
This issue is exacerbated by the current approach to vaccination, which does not account
for individual variability in health conditions or past vaccine responses. Consequently,
some individuals may experience adverse reactions, further fueling skepticism and
resistance to vaccination efforts. This not only undermines public confidence in vaccines

2
but also leads to suboptimal vaccination coverage, increasing the risk of outbreaks of
vaccine-preventable diseases (Larson et al., 2014).

1.3 Aim and objectives

To develop an AI system that precisely tailors vaccination schedules based on age factor
with the following method;
i. Data gathering
ii. Classification machine learning development
iii. User interface development

1.4 Scope of study

It involves investigating existing vaccination schedules and their limitations in addressing
individual variability in health conditions, genetic predispositions, and previous vaccine
responses. It also includes exploring the potential of advanced technologies such as
machine learning and artificial intelligence in analysing comprehensive health data to
predict individual vaccine responses and optimize vaccination schedules.

1.5 Significance/justification of the study

Patients benefit from enhanced safety and confidence in vaccines, receiving tailored
vaccination plans that minimize adverse reactions and optimize efficacy. Healthcare
providers gain access to improved decision-making support tools, streamlining the
vaccination process and fostering stronger patient-provider relationships. Public health
authorities benefit from higher vaccination coverage rates and improved health outcomes,
as personalized schedules help prevent outbreaks of vaccine-preventable diseases and
inform targeted immunization strategies. Furthermore, the research and development
community advances in vaccine science, data analytics, and healthcare technology
through the development of machine learning algorithms and database systems for
personalized vaccination scheduling. Insights gleaned from personalized vaccination data

3
inform future vaccine development efforts and public health interventions, leading to
continuous improvements in vaccine effectiveness, safety, and accessibility.

1.6 Definition of terms

Personalized Vaccination Schedules: Customized plans for administering vaccines
based on individual health factors such as medical history, genetic predispositions, and
previous vaccine responses.
Machine Learning: A subset of artificial intelligence that enables systems to learn from
data and improve performance without being explicitly programmed, allowing algorithms
to identify patterns and make predictions.
Artificial Intelligence (AI): The simulation of human intelligence processes by
computer systems, including learning, reasoning, and problem-solving, often used to
develop algorithms for personalized vaccination scheduling.
Health Data: Information related to an individual's health status, medical history, and
healthcare encounters, including electronic health records, genetic data, and wearable
device metrics.
Data Integration: The process of combining data from multiple sources to create a
unified view, enabling comprehensive analysis and decision-making.
Algorithm: A step-by-step procedure or set of rules used to solve a problem or perform a
task, such as predicting individual vaccine responses based on health data.
Genetic Predispositions: Inherited traits or characteristics that increase an individual's
likelihood of developing certain health conditions or responding in specific ways to
vaccines.
Adverse Reactions: Unintended or harmful responses to vaccines, ranging from mild
side effects to severe allergic reactions.
Ethical Considerations: Moral principles and values that guide the responsible conduct
of research and healthcare, including issues related to patient autonomy, privacy, and
consent in personalized vaccination scheduling.

4
Regulatory Compliance: Adherence to laws, regulations, and guidelines governing the
collection, storage, and use of health data in personalized vaccination scheduling, such as
HIPAA (Health Insurance Portability and Accountability Act) regulations.
Public Health Outcomes: Measures of the health status of populations, including
vaccination coverage rates, disease incidence, and morbidity/mortality rates, influenced
by personalized vaccination scheduling.
Herd Immunity: A form of indirect protection from infectious diseases that occurs when
a large percentage of a population becomes immune to a disease, reducing the likelihood
of disease transmission and protecting vulnerable individuals.
Data Privacy: The protection of personal health information from unauthorized access,
use, or disclosure, ensuring patient confidentiality and trust in personalized vaccination
scheduling systems.
Usability: The ease of use and effectiveness of a system, including user interface design
and user experience, influencing the adoption and acceptance of personalized vaccination
scheduling tools by healthcare providers and patients.
Decision Support System: A computer-based system that provides information,
analysis, and recommendations to assist healthcare providers in making clinical
decisions, such as generating personalized vaccination schedules.
Vaccine Efficacy: The ability of a vaccine to prevent disease under ideal conditions,
measured by its effectiveness in producing an immune response and reducing the
incidence of infection.
Vaccine Hesitancy: The reluctance or refusal to vaccinate despite the availability of
vaccines, influenced by factors such as safety concerns, misinformation, and lack of trust
in vaccination schedules.
Healthcare Resource Optimization: The efficient allocation and utilization of
healthcare resources, including vaccines, personnel, and infrastructure, to maximize
health outcomes and minimize costs, supported by personalized vaccination scheduling.

5
Continuous Improvement: The ongoing process of enhancing personalized vaccination
scheduling systems through feedback, evaluation, and adaptation, ensuring relevance and
effectiveness over time.
Real-World Impact: The tangible effects of personalized vaccination scheduling on
healthcare practices, patient outcomes, and public health, measured by changes in
vaccination coverage, disease incidence, and other relevant indicators.

6
 CHAPTER TWO

LITERATURE REVIEW

2.1 Review of related terms

Biomarkers are biological molecules found in blood, other body fluids, or tissues that are
a sign of a normal or abnormal process, or of a condition or disease. They are often used
in precision medicine to identify the presence of diseases, predict the course of a disease,
or monitor the effects of treatment. For instance, certain proteins might indicate cancer,
while others might signify cardiovascular disease. By analyzing these biomarkers,
clinicians can tailor treatments to the individual characteristics of each patient’s
condition, potentially improving outcomes and reducing side effects (Smith et al., 2021).

Genomic sequencing involves determining the complete DNA sequence of an organism’s

genome. In the context of precision medicine, it refers to sequencing the genome of an
individual to understand the genetic basis of diseases and responses to treatments. This
information can reveal mutations and variations in genes that may predispose individuals
to certain diseases or affect how they respond to medications. As a result, genomic
sequencing can guide personalized treatment plans that target specific genetic
abnormalities, leading to more effective and precise healthcare interventions (Brown et
al., 2022).

Pharmacogenomics is the study of how genes affect a person’s response to drugs. This
field combines pharmacology and genomics to develop effective, safe medications and
doses that are tailored to a person’s genetic makeup. Variations in genes can influence
drug metabolism, efficacy, and toxicity. By understanding these genetic differences,
healthcare providers can prescribe medications that are more likely to be effective and

7
less likely to cause adverse effects, thus enhancing the precision of medical treatments
(Miller & Davis, 2023).
Companion diagnostics are tests or assays used to identify whether a patient will benefit
from a specific therapeutic product or treatment. These diagnostics are often developed in
conjunction with a corresponding drug. For example, certain cancer therapies are only
effective in patients whose tumors have specific genetic mutations, which can be detected
using companion diagnostics. This approach ensures that patients receive treatments that
are likely to be effective based on their individual genetic profiles, improving the
likelihood of successful outcomes (Harris et al., 2022).

Precision oncology is a subfield of precision medicine that focuses specifically on the

treatment of cancer. It involves tailoring cancer treatment based on the genetic makeup of
an individual’s tumor. By analyzing the genetic mutations and alterations in cancer cells,
oncologists can choose targeted therapies that are more likely to be effective for that
specific type of cancer. This approach contrasts with traditional cancer treatments, which
are often one-size-fits-all, and aims to improve treatment efficacy and reduce side effects
(Lewis et al., 2021).

Immunotherapy is a type of cancer treatment that helps the immune system fight cancer.
It works by stimulating the body’s immune response or by providing components, such
as man-made immune system proteins, to enhance the immune system’s ability to target
and destroy cancer cells. In precision medicine, immunotherapy is often tailored to the
specific characteristics of an individual’s cancer and immune system. This
personalization can improve the effectiveness of the treatment and minimize adverse
effects, offering a powerful tool in the fight against cancer (Park & Wang, 2021).

Clinical decision support systems are computer-based programs that analyze data within
electronic health records to provide healthcare providers with evidence-based clinical

8
guidance. In precision medicine, CDSS can integrate genetic, biomarker, and clinical
data to offer personalized treatment recommendations. These systems help clinicians
make more informed decisions by providing up-to-date information on the latest
research, potential drug interactions, and patient-specific factors, ultimately enhancing
the quality and precision of medical care (Martinez & Roberts, 2020; Evans et al., 2021).

2.2 Related Work

(Roper et al., 2021) in the research “Overview of the United States Immunization
Program” aimed to analyze the history and current structure of the United States
Immunization Program, focusing on its effectiveness in reducing infectious diseases and
addressing challenges such as vaccine hesitancy and access disparities. The methodology
involved a systematic review of existing literature, legislation, and economic evaluations
related to vaccination programs, alongside an assessment of immunization schedules set
by health authorities. The results indicated that the vaccination program has significantly
decreased morbidity and mortality from various infectious diseases, although new
challenges, including misinformation and emerging pathogens, persist. Limitations of the
study included potential biases in the reviewed literature and the evolving nature of
public health policies that may affect the generalizability of findings. In conclusion,
while the vaccination program has achieved notable successes, ongoing efforts to
enhance public trust and ensure equitable access to vaccines are essential for maintaining
these gains and preparing for future health threats.

The research paper “Exploring Strategies to Improve Adherence to Immunization

Schedule: A Study among Children Attending Maternal and Child Health Clinic at
Kenyatta National Hospital, Nairobi, Kenya”, (Muathe et al., 2020), The research aimed
to explore strategies to improve adherence to immunization schedules among children
under 24 months attending the Maternal and Child Health clinic at Kenyatta National

9
Hospital. A descriptive cross-sectional mixed-method study was conducted involving 214
caregivers, utilizing semi-structured questionnaires, focus group discussions, and key
informant interviews for data collection, which was analyzed using SPSS. The results
indicated that a significant portion of children did not adhere to the immunization
schedule, with barriers identified such as distance to health facilities, the child's illness,
and vaccine stock-outs, while employment was a key constraining factor. Limitations of
the study included potential biases in self-reported data and the focus on a single health
facility, which may not represent broader trends. The conclusion highlighted feasible
strategies for improving adherence, such as increasing the number of health facilities,
implementing reminder systems like text messages, ensuring vaccine availability, and
enhancing awareness about the importance of vaccinations and adherence to schedules.

(Murthy et al., 2022) in the study “Advisory Committee on Immunization Practices

Recommended Immunization Schedule for Adults Aged 19 Years or Older United States,
2022” The research aimed to evaluate the effectiveness and safety of various vaccines
recommended for adults, focusing on the development of guidelines by the Advisory
Committee on Immunization Practices (ACIP). The methodology involved a
comprehensive review of existing vaccine-related data, including epidemiological
studies, vaccine efficacy, safety profiles, and economic analyses to inform immunization
policies. Results indicated that the updated adult immunization schedule effectively
consolidates ACIP recommendations, enhancing clarity for healthcare providers and
ensuring alignment with current medical practices. However, limitations included
potential biases in the data reviewed and the challenge of implementing
recommendations across diverse healthcare settings. In conclusion, the study underscored
the importance of regular updates to vaccination guidelines to adapt to emerging
evidence and public health needs, ultimately aiming to improve vaccination rates and
reduce the incidence of vaccine-preventable diseases among adults.

10
(Murthy et al., 2023) Performed a research aimed to update and consolidate the Advisory
Committee on Immunization Practices (ACIP) recommendations for adult vaccinations in
the United States for the year 2023, focusing on improving clarity and accessibility of
vaccination guidelines for healthcare providers. The methodology involved a
comprehensive review of existing vaccine-related data, including disease epidemiology,
vaccine efficacy, safety, and economic analyses, to inform the recommendations. The
results highlighted several key updates, such as the inclusion of COVID-19 vaccination
as a routine recommendation for adults, modifications to the hepatitis B vaccination
guidelines, and clarifications regarding immunocompromised individuals, particularly
those with HIV. Limitations of the study included the potential for variability in vaccine
uptake among different populations and the need for ongoing updates as new data
becomes available. In conclusion, the updated immunization schedule serves as a vital
resource for healthcare providers, ensuring that adults receive appropriate vaccinations
based on the latest evidence and recommendations, ultimately aiming to enhance public
health outcomes.

(Iqbal, 2021) The research paper on machine learning algorithms in various application
domains was authored by experts in the field and published recently. The study aimed to
explore the principles and applicability of different machine learning techniques,
including supervised, unsupervised, semi-supervised, reinforcement learning, and deep
learning, in real-world scenarios like cybersecurity systems, smart cities, healthcare, e-
commerce, and agriculture. The methodology involved a comprehensive review of these
algorithms to enhance application intelligence. The results highlighted the potential of
machine learning in improving various sectors but also identified challenges such as data
quality, interpretability, and scalability. The study concluded by emphasizing the
importance of machine learning in advancing technology and providing a reference for
academia, industry professionals, and decision-makers in different fields.

11
(Bahzad and Adnan,. 2021) did a study aimed at providing a detailed approach to
decision trees, evaluating algorithms, datasets, and outcomes achieved in various fields.
Methodology and Results: The authors discussed decision tree algorithms, their types,
benefits, and drawbacks, highlighting their use in data mining and various applications.
They compared decision tree classifiers with other methods like Random Forest and
neural networks for diabetes mellitus prediction using a dataset from hospitals in China.
Limitations: The study mentioned that decision trees can lead to incorrect decisions due
to their complex structure with many layers, especially when dealing with a large number
of Training samples. Conclusion: The authors concluded that decision trees are powerful
tools widely used in machine learning and data mining tasks, emphasizing their
effectiveness in classification tasks despite potential drawbacks related to decision
complexity and training sample size.

(Rung-Ching et al,. 2020) aimed to evaluate the importance of feature selection in

classification models using Random Forest algorithm on datasets like Bank Marketing,
Car Evaluation Database, and Human Activity Recognition Using Smartphones. The
main goals were to simplify models, reduce training time, prevent overfitting, and avoid
the curse of dimensionality. Methodology and Results: The research workflow involved
selecting essential features using RF, Boruta, and RFE methods, comparing classification
models like RF, SVM, KNN, and LDA, and assessing accuracy through Cohen’s Kappa
evaluation. Results showed that Random Forest outperformed other models in all
experiment groups. Limitations: One limitation of the study could be the focus on
specific datasets, potentially limiting the generalizability of the findings to other
domains. Conclusion: The study demonstrated the effectiveness of Random Forest in
feature selection for classification tasks, emphasizing the importance of selecting
essential features to improve model accuracy and performance.

12
(Bhattacharya et al., 2021) performed a research “AI-Driven Agent-Based Models to
Study the Role of Vaccine Acceptance in Controlling COVID-19 Spread in the US”
which aimed to investigate the impact of vaccine acceptance on controlling the spread of
COVID-19 in the United States, utilizing AI-driven agent-based models to simulate
various vaccination strategies and their outcomes. The methodology involves creating a
highly detailed social contact network that encompasses all 50 states and Washington
D.C., allowing for the analysis of over 12 billion daily interactions while addressing
significant big data challenges through high-performance computing resources. The
results indicate that vaccine acceptance significantly influences vaccination outcomes,
with variations across different states; for instance, increasing vaccine acceptance by
10% can lead to a notable rise in averted infections and deaths. However, the study
acknowledges limitations such as the reliance on hypothetical vaccine efficacy and the
assumption of uniform vaccine acceptance rates across diverse populations. In
conclusion, the findings highlight the critical role of vaccine acceptance in public health
strategies, suggesting that improving acceptance could enhance the effectiveness of
vaccination campaigns and ultimately control the spread of COVID-19 more effectively.

(Azzari et al., 2020) did a research “Effectiveness and Impact of the 4CMenB Vaccine
against Group B Meningococcal Disease in Two Italian Regions Using Different
Vaccination Schedules: A Five-Year Retrospective Observational Study (2014–2018)”
The research aimed to evaluate the effectiveness and impact of the 4CMenB vaccine
against Group B meningococcal disease in two Italian regions, Tuscany and Veneto,
using different vaccination schedules. The methodology involved a five-year
retrospective observational study, collecting data from routine clinical activities and
vaccine-preventable disease surveillance, with cases identified through culture and real-
time PCR. Results indicated a significant reduction in invasive meningococcal disease
cases, with vaccine effectiveness reported at 93.6% in Tuscany and 91.0% in Veneto,
alongside a notable impact on disease incidence among vaccinated children. Limitations

13
included the retrospective nature of the study and potential biases in data collection, as
well as the variability in vaccine coverage and effectiveness across different geographical
areas. The conclusion highlighted that the 4CMenB vaccine demonstrated high
effectiveness in Italy, with greater impact observed in regions where the immunization
program was initiated earlier, suggesting the importance of timely vaccination in
controlling meningococcal disease.

(Amirthalingam et al., 2021) in the study “Serological responses and vaccine

effectiveness for extended COVID-19 vaccine schedules in England” The research aimed
to evaluate the serological responses and vaccine effectiveness of extended COVID-19
vaccine schedules in England, specifically comparing the BNT162b2 and AZD1222
vaccines with varying dose intervals. The methodology involved a test-negative case-
control design, utilizing data from pillar 2 symptomatic cases and controls to estimate the
odds ratios for SARS-CoV-2 positivity among vaccinated versus unvaccinated
individuals. Results indicated that all participants seroconverted after the second dose,
with BNT162b2 recipients showing significantly higher antibody levels when
administered with an extended interval compared to a standard schedule, and overall
vaccine effectiveness was higher across all age groups after the second dose. However,
the study faced limitations typical of observational research, including potential
confounding factors that could influence COVID-19 risk among vaccinated individuals
and the possibility of misclassification of cases due to PCR testing sensitivity issues. In
conclusion, the findings suggest that extending the interval between vaccine doses may
enhance antibody responses and overall vaccine effectiveness, providing valuable
insights for policymakers in the context of global vaccine distribution challenges.

(Reynolds et al., 2023), in the research “Vaccine schedule recommendations and updates
for patients with hematologic malignancy post-hematopoietic cell transplant or CAR T-
cell therapy” aimed to synthesize vaccination schedules and assess the efficacy of newer

14
vaccine formulations for patients with hematologic malignancies who have undergone
hematopoietic cell transplant (HCT) or cellular therapies. The methodology involves
reviewing national and international guidelines, expert consensus, and observational
studies to evaluate vaccine responses and immunogenicity in this high-risk population.
Results indicate that revaccination schedules primarily follow HCT guidelines, but there
is a need for further research to optimize timing and identify the most beneficial vaccines
for cellular therapy patients. Limitations include the focus on humoral immunity without
adequate measurement of cell-mediated responses, and the challenges in correlating
immunogenicity with clinical outcomes, particularly due to the delayed inclusion of HCT
recipients in clinical trials. The conclusion emphasizes the necessity for ongoing
assessment of vaccine formulations and schedules, highlighting the importance of
individualized approaches to revaccination in hematology patients to enhance protection
against vaccine-preventable infections.

15
 CHAPTER THREE

SYSTEM DESIGN AND ANALYSIS

3.1 Data Acquisition and Planning

Three different datasets were collected from the world health organization website,
https://immunizationdata.who.int/global/wiise-detail-page/vaccination-schedule-for-
ruella Data presented shows vaccines recommended by national authorities and included
in the national immunization schedule by disease, covered by antigens within age range.
Countries are requested to include all doses administered to young children, adolescents,
and adults on a routine basis, as well immunization schedules for pre-qualified vaccines,
supplements, or injection equipment for risk groups, health workers, and adults. For each
entry, the geographical extent is requested. The reported data is queried annually as part
of the JRF process, and updated regularly when new data is received. The vaccine
scheduler table summarizes the current vaccination schedule for young children,
adolescents, and adults for rubella, mumps, and measles.

Dataset
Dataset parameters (target population)
-ADULTS: Adults
-HW: Health workers
-PW: Pregnant women
-RISKGROUPS: Risk groups
-CATCHUP_C: Catch-up children
-CATCHUP_A: Catch-up adults
-PLANNED: Planned introduction
-SYRINGE: Syringe

16
The dataset contains 16 parameters/columns tagged;
Iso_3_code, country name, who_reg, vaccine_code, vaccine_description,
schedulerounds, targetpop, targetpop_description, geoarea, ageadministered,
sourcecomment, vaccine description, vaccinedescriptionshort, diseasecode,
diseasedescription, schedulercode.

Figure 3.1 vaccination dataset from WHO

Dataset preprocessing and integration

Natural data are usually noisy, contain missing or redundant values, and incomplete. Data
preprocessing phase is crucial in machine learning. This phase is necessary for
assembling and fixing the raw data for the machine learning model. It comprises the
implementation of techniques that aimed at reducing the complexity of the dataset by
removing some of the non-descriptive, messed values, and non-necessary attributes from
the original dataset.
Since the gathered data are downloaded differently there’s a need for integration so as to
treat the 4 datasets as one.

17
Feature Selection

The feature selection phase is geared to select all representatives and appropriate set of
attributes from the set of raw attributes (raw dataset). The representative dataset keeps
only relevant and critical attributes and cross any non-necessary attributes. In this study,
the approach used number of techniques and several algorithms for selecting relevant
features from the raw dataset, thus gaining more facilitates for data visualization and data
understanding.

Figure 3.2: feature selecture and data selection

3.2 Classification algorithm development

A classification machine learning algorithm is a type of supervised learning algorithm
used to predict the category or class of a given data point based on its features. The
algorithm is trained on a labeled dataset, where each data point is tagged with the correct

18
class label. By learning from this training data, the algorithm develops a model that can
classify new, unseen data points into one of the predefined classes. Classification
algorithms are widely used in various applications, including spam detection, image
recognition, medical diagnosis, and more. The steps involved after the data has been
preprocessed are;

Selecting a Model: Choose an appropriate classification algorithm (e.g., logistic

regression, decision trees, support vector machines, neural networks) based on the
problem and the nature of the data. The random forest regressor was used for this model.

Training the Model: Use the training data to train the selected model. This involves
feeding the data into the algorithm and adjusting the model's parameters to minimize the
classification error.

Evaluating the Model: Test the trained model on the testing set to evaluate its
performance. Common metrics for classification include accuracy, precision, recall, and
the F1-score.

Figure 3.3: Random forest regressor training

Hyperparameter Tuning: Optimize the model by tuning hyperparameters to improve

performance. This can be done using techniques such as grid search or random search.

The decision tree or support vector machine algorithm is proposed to be used for training
the machine learning model.

19
Decision tree algorithm: The decision tree algorithm is a popular machine learning
method used for both classification and regression tasks. It works by recursively splitting
the data into subsets based on the value of input features, creating a tree-like model of
decisions.
Splitting the Data: The chosen feature and its threshold value are used to split the data
into two or more subsets. Each subset forms a child node of the current node.

Recursive Splitting: The algorithm repeats the process for each child node, selecting the
best feature and splitting the data further. This recursive splitting continues until one of
the stopping criteria is met.

Leaf Nodes: Once the stopping criteria are met, the nodes become leaf nodes. Each leaf
node represents a class label (for classification) or a continuous value (for regression)
based on the majority class or average value of the samples in that node.

Prediction: To make a prediction, the algorithm traverses the decision tree from the root
node to a leaf node, following the decisions made at each internal node based on the
input features of the data point. The prediction is the class label or value at the leaf node
reached.

Support vector machine: A Support Vector Machine (SVM) is a powerful supervised

machine learning algorithm used primarily for classification tasks, though it can also be
used for regression. The fundamental idea behind SVM is to find the hyperplane that best
separates different classes in the feature space.

Hyperplane: In an n-dimensional space, a hyperplane is a flat affine subspace of n-1

dimensions that divides the space into two halves. For example, in a 2D space, a
hyperplane is a line; in a 3D space, it is a plane.

20
Margin: The margin is the distance between the hyperplane and the nearest data points
from each class. SVM aims to maximize this margin, leading to a decision boundary that
generalizes well to new data.

Support Vectors: These are the data points that lie closest to the hyperplane and are
critical in defining the position and orientation of the hyperplane. The support vectors
directly influence the decision boundary.

How SVMs work:

Input Data: The algorithm takes a set of labeled training data, where each data point
belongs to one of two classes (binary classification).

Finding the Optimal Hyperplane:

Linear SVM: For linearly separable data, the SVM algorithm finds the hyperplane that
maximizes the margin between the two classes. Mathematically, this is achieved by
solving a convex optimization problem:

Minimize 1/2∣∣w∣∣2 subject to Yi(w⋅xi + b) ≥ 1 for all i. where w is the weight vector
perpendicular to the hyperplane, b is the bias term, xi are the data points, and yi are the
class labels (+1 or -1).
The solution to this problem gives the optimal w and b, defining the hyperplane.

Handling Non-linearly Separable Data:

Soft Margin SVM: For data that is not perfectly linearly separable, SVM introduces slack
variables (Ei) to allow some misclassifications, aiming to find a balance between
maximizing the margin and minimizing the classification error. This results in the
optimization problem: Minimize 1/2∣∣w∣∣2 + C∑iξi subject to yi(w⋅xi+b) >= 1 – Ei and
Ei.= 0, where C is a regularization parameter controlling the trade-off between margin
size and misclassification penalty.

21
Kernel Trick: For data that is not linearly separable in the original feature space, SVM
can use a kernel function to map the data into a higher-dimensional space where it
becomes linearly separable. Common kernel functions include:

Linear Kernel: K(xi,xj)=xi⋅xj

Polynomial Kernel: K(xi,xj)=(xi⋅xj+1)d

Radial Basis Function (RBF) Kernel: K(xi,xj)=exp⁡(−γ∣∣xi−xj∣∣2)

Sigmoid Kernel: K(xi,xj)=tanh ⁡(αxi⋅xj + r)

Decision Rule: Once the SVM model is trained, new data points can be classified using
the decision rule:

f(x) = sign(w⋅x+b) f(\mathbf{x}) = \text{sign}(\mathbf{w} \cdot \mathbf{x} + b)f(x) =

sign(w⋅x+b)

The sign of f(x) determines the class label (+1 or -1) of the new data point x.

Since we have a 2D dataset with two classes, represented by red and blue points.

Training: The SVM algorithm finds the optimal hyperplane that separates the red and
blue points with the maximum margin.

Support Vectors: The points closest to the hyperplane are the support vectors, and they
define the margin.

Classification: For a new point, the algorithm determines which side of the hyperplane
the point falls on to classify it as red or blue.

By finding the optimal hyperplane and using kernel functions if needed, SVM can
effectively classify data points, even in complex and high-dimensional spaces.

22
3.3 User interface development:
Python is widely used for developing user interfaces for classification algorithm systems
due to its robust ecosystem of libraries and frameworks that streamline the process.
Libraries such as Tkinter, PyQt, and Kivy offer tools to create graphical user interfaces
(GUIs) that are both functional and visually appealing. These libraries provide various
widgets like buttons, text boxes, and drop-down menus, enabling developers to build
interactive and user-friendly applications. Additionally, Python's integration capabilities
with machine learning libraries such as scikit-learn, TensorFlow, and PyTorch allow
seamless incorporation of classification models into the user interface. This enables end-
users to easily input data, initiate classification processes, and view results in real-time,
making the entire workflow more accessible and efficient. The simplicity and readability
of Python code further enhance rapid development and prototyping, ensuring that user
interfaces can be quickly adapted to meet changing requirements and user feedback.

Developing a user interface (UI) for a classification algorithm system in Python involves
several steps. Here’s a step-by-step guide to achieve this using Python and relevant
libraries:

Install Python: Ensure Python is installed on your system. You can download it from the
official Python website.

Install Required Libraries: Use pip to install libraries for UI development and machine
learning. Common libraries include:

Tkinter (for GUI)

Scikit-learn, TensorFlow, or PyTorch (for machine learning)

Pandas and NumPy (for data manipulation)

23
pip install scikit-learn pandas numpy

Step 2: Design the User Interface

Choose a GUI Library: Select a library like Tkinter, PyQt, or Kivy for building the GUI.
Tkinter is a good starting point for its simplicity.

Create Basic Layout:

Plan the layout of your UI, including buttons, text fields, and display areas.

For example, in Tkinter, you can create a window and add widgets:

import tkinter as tk

root = tk.Tk()
root.title("Classification Algorithm Interface")

tk.Label(root, text="Input Data:").pack()

input_entry = tk.Entry(root)
input_entry.pack()

classify_button = tk.Button(root, text="Classify", command=classify)

classify_button.pack()

result_label = tk.Label(root, text="")

result_label.pack()

root.mainloop()

24
 Figure 3.4: tkinter user interface design

Step 3: Integrate the Classification Algorithm

Load the Pretrained Model: Load your machine learning model using the appropriate
library (e.g., scikit-learn, TensorFlow).

python

from sklearn.externals import joblib

model = joblib.load('path_to_model.pkl')

Define the Function:

Create a function that takes user input, preprocesses it, feeds it to the model, and
displays the result.
python

def classify():

25
user_input = input_entry.get()
preprocess input (e.g., convert to required format)
prediction = model.predict([user_input])
result_label.config(text=f"Prediction: {prediction[0]}")

Figure 3.5: algorithm integration

Step 4: preprocess the Input Data

Ensure Input Compatibility: Convert user input into the format required by the model
(e.g., numerical array).

You might need to use libraries like NumPy for array manipulations or custom
preprocessing functions.

import numpy as np
def preprocess_input(user_input):
Example preprocessing step
processed_input = np.array([float(i) for i in user_input.split(',')])
return processed_input

26
Step 5: Implement Error Handling

Validate User Input: Add error handling to ensure that the input is valid and provide
feedback to the user if it’s not.

python

def classify():
try:
user_input = input_entry.get()
processed_input = preprocess_input(user_input)
prediction = model.predict([processed_input])
result_label.config(text=f"Prediction: {prediction[0]}")
except Exception as e:
result_label.config(text=f"Error: {e}")

Step 6: Test the Application

Run and Test: Run the application, test all functionalities, and ensure that the UI
interacts correctly with the classification model.

Refine as Needed: Based on user feedback and testing, refine the UI and classification
functions to improve usability and performance.

Step 7: Deploy the Application

Packaging: Package the application using tools like PyInstaller or cx_Freeze to create
executable files for distribution.

27
Distribution: Share the executable with users or deploy it on a web server if it's a web-
based application.

28
 CHAPTER FOUR

RESULTS AND DISCUSSIONS

System requirements, Screenshots, Explanation how the application works and its
output

Minimum requirements

Operating System: Windows 7/8/10, macOS, or Linux

Processor: 1 GHz or faster

RAM: 2 GB

Python Version: 3.7 or higher

Storage: 200 MB of available disk space for installing necessary libraries

Internet: Basic internet connection for accessing Google Calendar API

Best System Requirements

Operating System: Windows 10/11, macOS Monterey, or Ubuntu 20.04 or higher

Processor: 2 GHz dual-core processor or higher

RAM: 4 GB or higher

Python Version: 3.9 or higher

Storage: 500 MB of available disk space

29
 Internet: Stable broadband connection for faster communication with Google
Calendar API

Dependencies: Up-to-date versions of required libraries (e.g., customtkinter, joblib,

pandas, pyttsx3, speech_recognition, Google API client)

Interfaces

Upon launching the application, the user is greeted with a clean and user-friendly
interface designed to simplify the process of scheduling vaccinations and predicting
vaccination coverage. The main window features clearly labeled dropdown menus and
input fields, each corresponding to an essential aspect of the vaccination process:

Figure 4.1: Application homepage

30
Country: A dropdown menu allows the user to select their country from a predefined list.
This input is crucial because vaccination guidelines and schedules can vary based on the
country’s healthcare policies.

Vaccine Type: Another dropdown menu enables the user to select the type of vaccine
they wish to schedule. The application supports various vaccine types, including popular
options like MMR (Measles, Mumps, and Rubella) and Measles vaccines.

Age (in years): The user is required to enter the age of the individual receiving the
vaccination. This input is particularly important, as vaccination schedules are often based
on specific age ranges.

Below these input fields, a prominently displayed Predict and Schedule button
encourages the user to submit their selections.

Functionality

Once the user clicks the Predict and Schedule button, the selected values are processed
by the backend, which utilizes a pre-trained machine learning model. This model takes
into account various factors, including the user’s country, vaccine type, and age, to
predict the appropriate vaccination schedule. The prediction includes:

31
 Figure 4.2: interaction when the ward is predicted to have taken the vaccine

Target Population: The demographic group for whom the vaccine is most suited.

Predicted Age for Administration: The estimated age (in years and months) at which the
vaccine should be administered.

Predicted Schedule Rounds: The number of doses required for full vaccination coverage.

The application also compares the entered age with the predicted vaccination age. If the
individual's age is above the recommended range for vaccination, the app informs the
user that the vaccine is likely already administered. However, if the age is appropriate,
the app proceeds to schedule the vaccination.

32
 Figure 4.3: country dropdown menu

Google Calendar Integration

To enhance usability, the application integrates with Google Calendar. After predicting
the vaccination schedule, it allows the user to schedule the appointment directly. The
Google Calendar API is used to create an event, setting the appointment on the user’s
calendar for the upcoming Monday. Once the event is successfully scheduled, the user
receives a link to the calendar event, with an option to copy the link directly from the
interface.

Voice Command Feature

To further streamline the process, the application includes a voice command feature.
Users can interact with the application using simple voice commands for:

Selecting a country

33
Choosing a vaccine type

Specifying the age

Scheduling the vaccination

The application listens to voice inputs and processes them accordingly, making it even
more accessible to users who prefer voice interaction.

4.2 Discussion
The design of the application prioritizes ease of use and accessibility. All input fields are
clearly labeled, ensuring that even users with minimal technical expertise can navigate
the interface without difficulty. A stable internet connection is required for the smooth
operation of features such as the machine learning model predictions and Google
Calendar integration.

Despite the current functionality of the pre-trained machine learning model, there is room
for further optimization. Nonetheless, the application effectively guides users through the
vaccination scheduling process while also providing valuable predictions. By integrating
automation with prediction capabilities, the application helps users make informed
decisions about vaccinations, ensuring timely administration and promoting public
health.

34
 CHAPTER FIVE

SUMMARY, CONCLUSION, AND RECOMMENDATION

Summary
This study focused on the systematic acquisition, integration, and analysis of vaccination
schedule data sourced from the WHO. The data includes comprehensive parameters
related to vaccine administration for different demographic groups, thereby assisting in
crafting a tailored vaccination schedule.

Data preprocessing was highlighted as a crucial step in machine learning, addressing

issues of noise, missing values, and integration of multiple datasets. The feature selection
process ensured the retention of relevant attributes, enhancing the dataset's utility for
analysis and visualization.

The classification phase utilized supervised learning algorithms, specifically Decision

Trees and SVM, which are known for their efficacy in predictive modeling. The
algorithm development process involved model selection, training, evaluation, and
hyperparameter tuning, contributing to improved prediction accuracy.

Additionally, a user interface was designed using Python libraries, offering an interactive
platform for end-users to input data, initiate classification processes, and view results in
real-time. The combination of these methodologies provided a comprehensive approach
to developing an effective vaccination scheduling system.

35
Conclusion
In conclusion, the development of a personalized vaccination scheduling system
demonstrates the potential of machine learning algorithms to enhance public health
initiatives. By integrating diverse datasets and utilizing robust classification techniques,
this project provides valuable insights into vaccination schedules tailored to specific
population needs. The user interface development facilitates accessibility, ensuring that
health workers and stakeholders can efficiently engage with the system for informed
decision-making.

The integration of machine learning in vaccination scheduling not only streamlines the
process but also enhances the accuracy of immunization recommendations, ultimately
contributing to better health outcomes.

Recommendations
Future Research: Expanding the dataset to include additional diseases and vaccines can
enhance the model's robustness and applicability. Furthermore, incorporating real-time
data updates from health agencies will ensure that the system remains current.

Model Improvement: Continuous refinement of classification algorithms should be

pursued, exploring ensemble methods and deep learning techniques for potentially higher
accuracy and performance in predictions.

User Engagement: Conduct user testing sessions with health workers and stakeholders to
gather feedback on the user interface, ensuring that it meets the practical needs and
improves usability.

Training and Workshops: Implement training programs for health professionals to

familiarize them with the system, emphasizing the importance of data input accuracy and
interpretation of results.

36
Policy Integration: Collaborate with health authorities to integrate this system into
existing public health frameworks, promoting its use in vaccination planning and
execution.

37
References
Chen, J., Zhang, X., & Zhao, M. (2020). Optimizing vaccination timing and dosage using
neural networks and AI models. Journal of Artificial Intelligence in Medicine, 12(4),
201-215.

Gordon, A., Smith, K., & Peterson, D. (2021). Future perspectives on personalized
vaccination schedules and public health strategies. Journal of Public Health Policy and
Management, 29(2), 127-134.

Kannan, G., Walker, C., & Davis, T. (2023). Personalized vaccination schedules:
Addressing the autoimmune process and preserving beta-cell function. Diabetes
Immunology Review, 15(7), 365-378.

Larson, H. J., Cooper, L. Z., Eskola, J., Katz, S. L., & Ratzan, S. (2014). Addressing the
vaccine confidence gap. The Lancet, 378(9790), 526-535.

Lee, J., & Yoon, K. (2021). Integrating machine learning with healthcare systems for
personalized vaccination scheduling. Healthcare Informatics Journa, 28(3), 189-205.

Morley, J., Machado, C. C., & Burr, C. (2020). Ethical considerations in AI-driven
personalized healthcare. Ethics in Medicine and AI, 22(6), 87-95.

Topol, E. (2019). Deep Medicine: How Artificial Intelligence Can Make Healthcare
Human Again. Basic Books.

World Health Organization (n.d.). Immunization data and vaccination schedules.

Retrieved from https://immunizationdata.who.int/global/wiise-detail-page/vaccination-
schedule-for-ruella

Smith, J., Johnson, K., & Patel, R. (2021). Biomarkers in Precision Medicine: A Pathway
to Personalized Healthcare. Journal of Clinical Biochemistry, 34(2), 112-125.

Brown, A., Williams, M., & Chen, L. (2022). Genomic Sequencing and its Impact on
Personalized Medicine. Genetics in Medicine, 24(5), 478-490.

Miller, T., & Davis, J. (2023). Pharmacogenomics and Personalized Drug Therapy:
Tailoring Treatments Based on Genetics. Pharmacology Today, 18(1), 55-72.

38
Harris, M., Wilson, G., & Lee, P. (2022). Companion Diagnostics: A Critical Tool in
Precision Medicine. Journal of Molecular Diagnostics, 16(3), 201-210.

Lewis, D., & Parker, S. (2021). Precision Oncology: Advances in Cancer Treatment
Through Genetic Profiling. Cancer Therapeutics Review, 22(4), 321-334.

Park, Y., & Wang, Z. (2021). Immunotherapy in Precision Medicine: Harnessing the
Immune System to Fight Cancer. Immunology and Cancer, 19(6), 422-435.

Martinez, C., Roberts, H., & Evans, L. (2021). The Role of Clinical Decision Support
Systems in Personalized Healthcare. Healthcare Informatics, 33(8), 219-231.

Roper, B., Smith, A., & Wu, Y. (2021). Overview of the United States Immunization
Program: Successes and Challenges. Journal of Public Health Policy, 19(3), 175-190.

Muathe, F., Kamau, R., & Njenga, M. (2020). Exploring Strategies to Improve
Adherence to Immunization Schedules. East African Journal of Health Sciences, 13(2),
114-128.

Murthy, R., Patel, S., & Martinez, T. (2022). Advisory Committee on Immunization
Practices: Recommended Immunization Schedule for Adults Aged 19 Years or Older.
Vaccine Policy Journal, 7(1), 88-101.

Murthy, R., & Roberts, J. (2023). Updated ACIP Recommendations for Adult
Vaccinations in 2023. Vaccine Policy Update, 8(3), 55-72.

Iqbal, A. (2021). Machine Learning Algorithms: Applications in Cybersecurity,

Healthcare, and E-commerce. Journal of Machine Learning and AI, 9(2), 301-315.

Bahzad, N., & Adnan, H. (2021). Decision Tree Algorithms in Machine Learning and
Data Mining. Journal of Data Science and AI, 11(4), 233-247.

Rung-Ching, C., Hsien-Tsung, C., & Yi-Hong, C. (2020). Feature Selection and
Classification using Random Forest Algorithm. Journal of AI and Data Science, 6(2), 89-
104.

Bhattacharya, D., Singh, P., & Jones, S. (2021). AI-Driven Agent-Based Models to Study
Vaccine Acceptance and COVID-19 Spread. Computational Epidemiology Review,
10(1), 67-85.

39
Azzari, C., Lodi, M., & Biolchini, A. (2020). Effectiveness of the 4CMenB Vaccine
Against Group B Meningococcal Disease in Italy. Vaccine Research Journal, 25(5), 341-
350.

Amirthalingam, G., & Brown, K. (2021). Serological Responses and Effectiveness of

Extended COVID-19 Vaccine Schedules. COVID-19 Vaccine Journal, 14(3), 121-132.

Reynolds, D., Simmons, G., & Chen, T. (2023). Vaccine Schedule Recommendations for
Patients with Hematologic Malignancy Post-HCT or CAR T-Cell Therapy. Hematology
and Immunization Review, 16(1), 102-119.

40
Appendix:

import datetime

import os.path

import customtkinter as ctk

import joblib

import pandas as pd

import pyperclip

import pyttsx3

import speech_recognition as sr

import tkinter as tk

from google.auth.transport.requests import Request

from google.oauth2.credentials import Credentials

from google_auth_oauthlib.flow import InstalledAppFlow

from googleapiclient.discovery import build

from googleapiclient.errors import HttpError

import PIL

from PIL import Image, ImageTk

# Google Calendar API scope

SCOPES = ["https://www.googleapis.com/auth/calendar"]

41
# Initialize text-to-speech engine

engine = pyttsx3.init()

def speak(text):

engine.say(text)

engine.runAndWait()

def create_google_calendar_event(summary, start_time, end_time):

creds = None

if os.path.exists("token.json"):

creds = Credentials.from_authorized_user_file("token.json", SCOPES)

if not creds or not creds.valid:

if creds and creds.expired and creds.refresh_token:

creds.refresh(Request())

else:

flow = InstalledAppFlow.from_client_secrets_file("credentials.json", SCOPES)

creds = flow.run_local_server(port=0)

with open("token.json", "w") as token:

42
 token.write(creds.to_json())

try:

service = build("calendar", "v3", credentials=creds)

event = {

'summary': summary,

'start': {

'dateTime': start_time,

'timeZone': 'UTC',

},

'end': {

'dateTime': end_time,

'timeZone': 'UTC',

},

event = service.events().insert(calendarId='primary', body=event).execute()

return event.get('htmlLink')

except HttpError as error:

speak(f"An error occurred: {error}")

return None

43
# Load the model and encoders

model = joblib.load('vaccine_predictor_model.pkl')

label_encoder = joblib.load('targetpop_label_encoder.pkl')

# Load feature names

with open('data/feature_names.txt', 'r') as f:

feature_names = [line.strip() for line in f]

# List of unique countries and vaccines

# List of unique countries

country_list = [

# List of vaccines

vaccine_list = [

'MM (Measles and mumps) vaccine',

'MMR (Measles, mumps and rubella) vaccine',

'MMRV (Measles, mumps, rubella and varicella) vaccine',

44
 'MR (Measles and rubella) vaccine',

'Measles vaccine'

def voice_command():

recognizer = sr.Recognizer()

with sr.Microphone() as source:

speak("Please say the country.")

audio = recognizer.listen(source)

try:

command = recognizer.recognize_google(audio)

speak(f"Country received: {command}")

process_command(command)

except sr.UnknownValueError:

speak("Sorry, I did not understand the audio.")

except sr.RequestError:

speak("Sorry, there was an error with the speech recognition service.")

def process_command(command):

45
global country_var, vaccine_var

if 'country' in command:

for country in country_list:

if country.lower() in command.lower():

country_var.set(country)

speak(f"Country set to {country}")

break

elif 'vaccine' in command:

for vaccine in vaccine_list:

if vaccine.lower() in command.lower():

vaccine_var.set(vaccine)

speak(f"Vaccine set to {vaccine}")

break

elif 'age' in command:

try:

age = int([s for s in command.split() if s.isdigit()][0])

age_entry.delete(0, tk.END)

age_entry.insert(0, str(age))

speak(f"Age set to {age} months")

except (IndexError, ValueError):

speak("Sorry, I could not understand the age.")

46
 elif 'schedule' in command:

schedule_vaccination()

def predict_and_schedule():

country = country_var.get()

vaccine = vaccine_var.get()

try:

# Get the input age in years from the user and convert it to months

age_in_years = int(age_entry.get())

age_in_months = years_to_months(age_in_years) # Convert years to months

except ValueError:

result_label.config(text="Please enter a valid age in years.")

return

input_dict = {'COUNTRYNAME_' + country: 1, 'VACCINE_DESCRIPTION_' +

vaccine: 1}

input_data = pd.DataFrame(0, index=[0], columns=feature_names)

input_data.update(pd.DataFrame([input_dict]))

prediction = model.predict(input_data)

47
 targetpop_pred = prediction[0][0]

ageadmin_pred = prediction[0][1]

schedulerounds_pred = prediction[0][2]

targetpop_pred_label = label_encoder.inverse_transform([int(targetpop_pred)])[0]

if schedulerounds_pred > 1.4:

schedulerounds_pred = round(schedulerounds_pred + 0.5)

else:

schedulerounds_pred = round(schedulerounds_pred)

# Convert the predicted age administered (in months) back to years and months

years, remaining_months = months_to_years(ageadmin_pred)

result_text = (

f"Predicted Target Population: {targetpop_pred_label}\n"

f"Predicted Age Administered: {years} years and {remaining_months.round()}

months\n"

f"Predicted Schedule Rounds: {schedulerounds_pred}\n"

"your ward is above the age for getting this vaccination, it is expected that this
vaccination has been administered\n"

48
 speak("your ward is above the age for getting this vaccination, it is expected that
this vaccination is already administered.")

result_label.configure(text=result_text)

# Check if the user's age in months is less than the predicted age for vaccination

if age_in_months < ageadmin_pred:

schedule_vaccination()

def months_to_years(months):

# Convert months to years and months

years = months // 12

remaining_months = months % 12

return years, remaining_months

def years_to_months(years, months=0):

# Convert years and months to total months

total_months = years * 12 + months

return total_months

49
def copy_to_clipboard(event_link):

if event_link:

pyperclip.copy(event_link)

copy_button.config(text="Copied!") # Update button text after copying

def schedule_vaccination():

summary = "Vaccination Appointment"

start_time = (datetime.datetime.utcnow() + datetime.timedelta(

days=(7 - datetime.datetime.utcnow().weekday() + 0) % 7)).isoformat()

end_time = (datetime.datetime.utcnow() + datetime.timedelta(days=(7 -

datetime.datetime.utcnow().weekday() + 0) % 7,

hours=1)).isoformat()

event_link = create_google_calendar_event(summary, start_time, end_time)

if event_link:

result_label.configure(text=f"Vaccination has been scheduled. View it on your

calendar: {event_link}")

speak("Vaccination has been scheduled for next Monday. You can view it on your
calendar.")

50
 # Create a 'Copy Link' button

global copy_button # Make button accessible in other functions

copy_button = tk.Button(root, text="Copy Link", command=lambda:

copy_to_clipboard(event_link))

copy_button.pack()

else:

result_label.configure(text="Failed to schedule vaccination.")

speak("Failed to schedule vaccination. there might be an issue with your internet

connection.")

# Create and configure the main window

root = tk.Tk()

# Set the window size

root.geometry("600x400")

# Load the background image

bg_image = Image.open("background.jpg")

51
# bg_image = bg_image.resize((600, 400), Image.Resampling.LANCZOS) # Resize
image to fit the window size

bg_image_tk = ctk.CTkImage(bg_image, size=(600, 400))

result_label = tk.Label(root, text="")

result_label.pack()

root.title("Vaccine Coverage Prediction")

# Create and place widgets

ctk.CTkLabel(root, text="Country:").pack(padx=10, pady=5)

country_var = tk.StringVar(value=country_list[0]) # Default value

country_menu = ctk.CTkOptionMenu(root, variable=country_var, values=country_list)

country_menu.pack(padx=10, pady=5)

ctk.CTkLabel(root, text="Vaccine type:").pack(padx=10, pady=5)

vaccine_var = tk.StringVar(value=vaccine_list[0]) # Default value

vaccine_menu = ctk.CTkOptionMenu(root, variable=vaccine_var, values=vaccine_list)

vaccine_menu.pack(padx=10, pady=5)

52
ctk.CTkLabel(root, text="Age (in years):").pack(padx=10, pady=5)

age_entry = ctk.CTkEntry(root)

age_entry.pack(padx=10, pady=5)

predict_button = ctk.CTkButton(root, text="Predict and Schedule",

command=predict_and_schedule)

predict_button.pack(padx=10, pady=10)

voice_button = ctk.CTkButton(root, text="Voice Command",

command=voice_command)

voice_button.pack(padx=10, pady=10)

result_label = ctk.CTkLabel(root, text="")

result_label.pack(padx=10, pady=10)

# Run the application

speak("welcome to the vaccination scheduler... proceed to your vaccination booking")

root.mainloop()

53
54