DEGREE: MSc Data Analytics

Module: Predictive Analytics and Machine Learning using Python

Assignment Title: Leveraging Random Forest, Regression, and KNN for Predicting
Cryptocurrency Prices
Assignment Type: Report
Word Limit: 2000-3000 words (+/- 250)
Weighting: 50%
Issue Date: 21/11/2024
Submission Date: 1 3 / 0 2 / 2 0 2 5
Feedback Date: 0 6 / 0 3 / 2 0 2 5
Plagiarism:
When submitting work for assessment, students should be aware of the InterActive/Canvas guidance
and regulations in concerning plagiarism. All submissions should be your own, original work.
You must submit an electronic copy of your work. Your submission will be electronically
checked.

Learner declaration

I certify that the work submitted for this assignment is my own and research sources
are fully acknowledged.

Student signature: Date:

Harvard Referencing:
The Harvard Referencing System must be used. The Wikipedia, UKEssays.com or similar websites
must not be used or referenced in your work.
 Introduction

Learning Outcomes:

LO1. Demonstrate the understanding of basic concepts of dealing with different types of data – ordinal, categorical,
encoding along with collecting, storing and making it ready for processing.

LO2. Explain the various components of predictive analytics, with the models for regression, classification and
clustering to analyze real-life business problems.

LO3. Implement various models and work on a project life cycle from end to end to solve an analytical problem
which translates into a business problem solution using machine learning and artificial intelligence.

Assessment Criteria: Weighting 50%

2000-3000 words

This assignment aligns with the unit's objectives of introducing learners to predictive analytics, machine learning,
and artificial intelligence (AI) concepts. You will apply these techniques using Python to solve a real-world problem:
predicting Bitcoin prices.

Objectives:

• Understand the concepts of feature selection and its application in machine learning for price prediction.
• Transform orderbook data into time-series features and utilize these features for model training.
• Implement and compare different machine learning models for Bitcoin price prediction.

Dataset:

For this assignment, you will use the provided dataset, which can be accessed at the following link:
https://github.com/usmanakhtar/Cryptocurrency-Dataset/

Phase 1: Data Exploration and Preparation (LO1, LO2, LO3)

• 1.1 Data Exploration (LO1):

o Identify relevant features for predicting Bitcoin prices.
o Visualize the price data over time series to identify trends, seasonality, and potential outliers.
o Calculate descriptive statistics for numerical features (mean, median, standard deviation) and
analyze categorical features (frequency distribution).
o Check for missing values and implement appropriate strategies (e.g., imputation, removal) to
handle them.

• 1.2 Data Preprocessing (LO2):

o Define the target variable as the future Bitcoin and Ethereum prices (e.g., closing price in the next
time window).
o Select relevant features based on the exploration phase. Consider feature engineering techniques
to create new features from existing ones (e.g., price change, price spread).
o Scale the features to a common range (e.g., standardization, normalization) to improve model
performance.
Phase 2: Feature Selection (LO2):

• 2.1 Feature Selection Techniques (LO2):

o Implement feature selection methods like:
▪ Filter methods: Select features based on statistical measures like correlation with the target
variable (e.g., Information Gain, chi-square test).
▪ Wrapper methods: Train multiple models with different feature subsets and choose the
one with the best performance (e.g., Recursive Feature Elimination (RFE)).
▪ Embedded methods: Feature selection is integrated within the model training process (e.g.,
LASSO regression).
o Compare the results of different feature selection methods and choose the most relevant features
for further modelling.

Phase 3: Model Training and Evaluation (LO2, LO3)

• 3.1 Model Selection and Training (LO2, LO3):

o Implement Random Forest, Regression, and KNN models for Bitcoin price prediction.
o Split the data into training and testing sets.
o Train each model on the training set.

• 3.2 Model Evaluation (LO2):

o Evaluate the performance of each model on the testing set using Root Mean Squared Error (RMSE).
o RMSE measures the average difference between predicted and actual prices.
o Compare the RMSE of different models and select the one with the lowest error for optimal price
prediction.

Phase 4: Conclusion and Future Work (LO3)

• Summarize the findings:

o Compare the performance of Random Forest, Regression, and KNN models.
o Discuss the impact of feature selection on model accuracy.
o Analyze the strengths and weaknesses of each model.
• Discuss potential limitations of the study:
o Address the inherent volatility of Bitcoin prices and limitations of historical data predicting future
trends.
• Suggest future work:
o Explore advanced feature engineering or deep learning models for potentially improved prediction
accuracy.
o Consider incorporating external factors like market sentiment or news analysis for a more
comprehensive model.
Deliverables:
• A Jupyter Notebook or Python script documenting your code for data exploration, preprocessing, feature
selection, model training, evaluation, and visualization of results.
• A report in BSBI template format summarizing your findings, including visualizations, chosen features,
model performance, limitations, and future work suggestions.
• The Harvard Referencing System must be used.
• Your essay should be 2000-3000 (+/- 250) words, excluding Title, Table of Contents, Bibliography and
Appendices.
• Please upload your submission as a single file (PDF or Doc) on the BSBI portal.
GRADING DESCRIPTORS: LEVEL 7

EXPERIMENTATION & INNOVATION

FAIL PASS
Threshold Criteria 0-29% 30-39% 40-49% 50-59% 60-69% 70-79% 80-89% 90-100%
Deals with complex Little to no ability to Low Limited Competent Good Very Good Excellent range of Exceptional
issues both use techniques to utilisation of research or understanding of understanding of problem-solving extremely well- problem-solving skills
systematically and deal with complex established advanced solving problems, solving problems skills displaying a developed problem- with sophisticated
creatively issues systematically techniques to deal scholarship to through own through own comprehensive solving displaying evaluation and
demonstrating self- (including those of with complex issues their area of research or research and understanding of an understanding application of a wide
direction and ethics and systematically study by using a advanced advanced techniques of techniques range of advanced
originality in tackling sustainability) and (including those of range of scholarship scholarship applicable to their applicable to their information and
and solving problems creatively to solve ethics and information and displaying a critically selecting own research or own research or techniques to
problems and/or sustainability) and established and comprehensive and displaying a advanced advanced undertake projects.
make decisions. creatively to solve advanced understanding of comprehensive scholarship scholarship beyond
problems and/or techniques established and understanding of which is taught.
make decisions, but advanced established and
with limitations in techniques advanced
techniques or techniques.
approach.
Comprehensive Little to no Low Limited Competent Good understanding Very good Excellent Exceptional
understanding of understanding of understanding of understanding of understanding of of techniques understanding of understanding of understanding of
techniques techniques applicable techniques key techniques techniques applicable applicable to their techniques applicable techniques applicable techniques
applicable to their to their own research applicable to their applicable to their to their own research own research or to their own research to their own research applicable to their
own research
own research or or advanced or advanced own research or or advanced advanced scholarship or advanced or advanced own research or
advanced scholarship or their scholarship including advanced scholarship including and a some scholarship and a scholarship and advanced
scholarship limitations and their limitations and scholarship including their limitations and understanding of some understanding mastery of some scholarship and
ambiguities. ambiguities. their limitations and ambiguities more specialised of more specialised more specialised mastery of some
ambiguities. techniques. techniques. areas. more specialised
areas.
GRADING DESCRIPTORS: LEVEL 7
RESEARCH & ANALYSIS
FAIL PASS
Threshold Criteria 0-29% 30-39% 40-49% 50-59% 60-69% 70-79% 80-89% 90-100%
Systematic Little to no Low knowledge of Limited knowledge to Competent Good knowledge of Very good knowledge Excellent knowledge Exceptional
understanding of knowledge of the the subject lacking deal with terminology, knowledge of ideas ideas or arguments at of ideas or arguments of ideas or arguments knowledge of ideas
knowledge, and a subject with limited coherence, breadth, facts and concepts or arguments at the the forefront of any at the forefront of at the forefront of or arguments at the
critical awareness of breadth or depth or or detail with only some of which is forefront of any part of the subject the subject some of the subject many of forefront of the
current problems deficiencies in major some reference to informed by the part of the subject showing a clear, which are which are subject most of which
and/or new insights, areas or currency. ideas or arguments at forefront of defined sufficient to deal critical insight into significantly beyond significantly beyond are significantly
much of which is at, the forefront of any areas of the subject. with current issues the discipline as what has been taught what has been taught beyond what has
or informed by, the part of the subject. in the discipline, whole and current and show a critical and show a critical been taught and
forefront of their generally more issues/problems. insight into the insight into the show a critical insight
academic discipline, descriptive than discipline and current discipline and current into the discipline
field of study or area critical or issues/problems. issues/problems. and current
of professional analytical. issues/problems.
practice

Conceptual Little to no conceptual Low conceptual Limited conceptual Competent Good conceptual Very good conceptual Excellent conceptual Exceptional
understanding that understanding or understanding and understanding and conceptual understanding which understanding which understanding which conceptual
enables the student argument and a focus arguments are weak argument understanding and critically evaluate and systematically critically apply a wide understanding of
to display originality on descriptive or poorly construction with argument synthesise other synthesises a wide range of views publishable quality
in the application of explanations which constructed, and the critical evaluation of construction with views and range of views with a through a perceptive with systematic
knowledge do not comment on work does not alternative views or critical evaluation information with a critical insight into use of advanced engagement and
arguments of others critically evaluate the comment on advanced of a range of views thoughtful advanced scholarship. usage of advanced
or alternative views. arguments of others scholarship. and consistent interpretation of scholarship. scholarship.
or consider engagement with advanced
alternative views. advanced scholarship.
scholarship.
GRADING DESCRIPTORS: LEVEL 7

ENGAGING WITH PRACTICE

FAIL PASS
Threshold Criteria 0-29% 30-39% 40-49% 50-59% 60-69% 70-79% 80-89% 90-100%
Practical Little to no evidence Low evidence of Limited Competent Good background Very good, Excellent Exceptional
understanding of of background background background investigation, investigation, independent, independent, investigation,
how established investigation, investigation, investigation, analysis, research, analysis, research, extensive and extensive and analysis, research,
techniques of analysis, research, analysis, research, analysis, research, enquiry, ethical enquiry, ethical appropriate appropriate enquiry, ethical
research and enquiry, ethical enquiry, ethical enquiry , ethical awareness, and/or awareness, and/or investigation, investigation, awareness, and/or
enquiry are used awareness, and/or awareness, and/or awareness, and/or study using study using analysis, research, analysis, research, study which
to create and study. study. study using established established enquiry, ethical enquiry, ethical demonstrates
interpret established techniques techniques awareness, and/or awareness, and/or carefully considered
knowledge in the techniques, with accurately, and can accurately, and study beyond the study well beyond depth and breadth
discipline the ability to critically appraise and possesses a well- usual range, and the usual range, and and critically
extract relevant use academic developed ability to critically evaluates critically evaluates synthesises this to
points. sources. critically appraise a this to advance the this to advance the advance the work
wide range of work and/or direct work and/or direct and/or direct
sources. arguments. arguments. arguments.

Originality in Little to no technical, Low technical, Limited Competent technical, Good technical, Very good range of Excellent range of Exceptional range of
the creative or artistic creative or artistic technical, creative or artistic creative or artistic technical, creative or technical, creative or technical, creative or
application of skills related to their skills related to their creative or skills required for skills required for artistic skills. artistic skills artistic skills
knowledge area of study. area of study. artistic skills area of study. area of study.
required for
area of study.
Independently Little to no Low contribution to Limited contribution Competent Good contribution to Very good Excellent Exceptional
advance your own contribution to group group activity and/or to group activity contribution to group group activity and/or contribution to group contribution to group contribution to group
knowledge and activity and/or undertaking further and/or undertaking activity and/or independently activity and/or activity and/or activity and/or
understanding, and undertaking further training at a further training at a independently undertakes further independently independently independently
to develop new training at a high/advanced level. high/advanced level. undertakes further training at a undertakes further undertakes further undertakes further
skills to a high high/advanced level. training at a high/advanced level training at a training at a training at a
level. high/advanced level. with an high/advanced level high/advanced level high/advanced level
understanding of with an with teamwork and with teamwork and
team roles understanding of leadership strong leadership.
team roles
GRADING DESCRIPTORS: LEVEL 7

REALISATION & COMMUNICATION

FAIL PASS
Threshold Criteria 0-29% 30-39% 40-49% 50-59% 60-69% 70-79% 80-89% 90-100%
Communicate Little to no clarity in Low clarity in the Limited clarity in the Competent Good, confident and Very good, confident Excellent Exceptional
information, ideas, the communication communication of communication of communication of clear communication and clear communication of communication of
problems and of ideas, problems ideas, problems and ideas, problems and ideas, problems and of ideas, problems communication of ideas, problems and ideas, problems and
solutions to both and solutions to solutions to solutions to solutions to and solutions to ideas, problems and solutions to solutions to
specialist and non- audiences. audiences. audiences. audiences. audiences in a range solutions to audiences in a range audiences in a range
specialist audiences. of means / media. audiences in a range of means / media. of means / media.
of means / media.
GRADING DESCRIPTORS: LEVEL 7

PERSONAL & PROFESSIONAL CONNECTIVITY

FAIL PASS
Threshold 0-29% 30-39% 40-49% 50-59% 60-69% 70-79% 80-89% 90-100%
Criteria
Independently Little to no Low contribution Limited Competent contribution to Good contribution Very good Excellent contribution Exceptional contribution to
advance your contribution to group activity contribution to group activity and/or to group activity contribution to to group activity group activity and/or
own knowledge to group and/or group activity independently undertakes and/or group activity and/or independently independently undertakes
and activity undertaking and/or further training at a independently and/or undertakes further further training at a
understanding, and/or further training at undertaking high/advanced level. undertakes further independently training at a high/advanced level with
and develop undertaking a high/advanced further training at training at a undertakes high/advanced level teamwork and strong
new skills to a further level. a high/advanced high/advanced further training with teamwork and leadership.
high level. training at a level. level with an at a leadership
high/advance understanding of high/advanced
d level. team roles level with an
understanding
of team roles
Qualities and Little to no Low ability to manage Limited ability to Competent ability to Good ability to Very good Excellent ability to Exceptional ability to
transferable ability to learning and/or manage learning manage learning, and systematically ability to manage learning on manage learning on own
skills manage exercise initiative, and exercise exercise initiative, manage systematically own initiative, and initiative, and exercise
necessary for learning and/or ethical and personal initiative, ethical ethical and personal learning, and manage exercise initiative, initiative, ethical and
employment exercise responsibility and/or and personal responsibility, and exercise learning, and ethical and personal personal responsibility,
requiring: (a) initiative, ethical decision-making in responsibility, decision-making in initiative, ethical exercise responsibility, and and decision-making in
the exercise and personal complex and and decision- complex and and personal initiative, ethical decision-making in complex and
of initiative, responsibility unpredictable situations making in unpredictable responsibility, and personal complex and unpredictable situations
ethical and and/or decision- complex and situations and decision- responsibility, unpredictable situations
personal making in unpredictable making in and decision-
responsibility complex and situations complex and making in
(b) decision- unpredictable unpredictable complex and
making in situations situations unpredictable
complex and situations.
unpredictable Little to no use Low use of appropriate Limited Competent expression, Good Very good Excellent expression, Exceptional expression,
contexts of appropriate terminology, with many expression, style style, and appropriate expression, expression, style style and appropriate style and appropriate
terminology, errors in spelling, and appropriate vocabulary with some style and and appropriate vocabulary with vocabulary with no
limited vocabulary and syntax. vocabulary with errors in spelling, appropriate vocabulary with minimal errors in errors in spelling,
vocabulary and errors in spelling, grammar and syntax vocabulary with minimal errors in spelling, grammar and grammar and syntax.
many errors in grammar and which do not affect some errors in spelling, grammar syntax.
spelling, syntax which understanding. spelling, and syntax.
grammar and affect grammar and
syntax. understanding. syntax.
GRADING DESCRIPTORS: LEVEL 7
Little to no Low evidence of basic Limited evidence Adequate evidence of Good evidence of Very good Excellent evidence of Exceptional evidence of
evidence of basic numeracy or digital of numeracy or numeracy or digital numeracy or digital evidence of numeracy or digital numeracy or digital
numeracy or literacy, hardware and digital literacy, literacy, hardware and literacy, hardware numeracy or literacy, hardware and literacy, hardware and
digital literacy, software skills hardware and software skills and software skills digital software skills software skills
hardware and competency. software skills competency. competency. literacy, competency. competency.
software skills competency. hardware
and
software
skills
competency. competency.
Does not demonstrate achievement of professional The student has demonstrated achievement of professional competence when assessed against the requirements of a PSRB.
competence when assessed against the requirements of a
professional, statutory or regulatory body (PSRB).
Inaccurate use of terminology with limited vocabulary and The student has adhered to the appropriate rules and/or conventions set by regulators or the industry.
many errors in spelling, grammar and syntax.
Inaccurate terminology, with many errors in spelling,
vocabulary and syntax.