Higher Nationals

Internal verification of assessment decisions – BTEC (RQF)

INTERNAL VERIFICATION – ASSESSMENT DECISIONS

Programme title BITEC Higher National Diploma in Computing

Assessor Internal Verifier

Unit 17 - Business Process Support

Unit(s)

Business Process of a Company.

Assignment title

Student’s name

List which assessment Pass Merit Distinction

criteria the Assessor has
awarded.

INTERNAL VERIFIER CHECKLIST

Do the assessment criteria awarded

match those shown in the assignment
brief? Y/N

Is the Pass/Merit/Distinction grade

awarded justified by the assessor’s Y/N

comments on the student work?

Has the work been assessed

accurately? Y/N

Is the feedback to the student:

Give details:
• Constructive?
• Linked to relevant assessment Y/N

criteria? Y/N

• Identifying opportunities for Y/N

improved performance?
 • Agreeing actions? Y/N

Does the assessment decision need

amending? Y/N

Assessor signature Date

Internal Verifier signature Date

Programme Leader signature (if

required) Date

Confirmation completed

Remedial action taken

Give details:

Assessor signature Date

Internal
Verifier Date

signature

Programme Leader
signature (if Date

required)
Assignment Feedback Form

Student Name/ID

Unit Title Unit 17 - Business Process Support

Assignment Number Assessor

Date
Submission Date Received 1st
submission

Date Received 2nd

Re-submission Date submission
Assessor Feedback:
LO1 Discuss the use of data and information to support business processes and the value they have
for an identified organization
Pass, Merit & Distinction P1 P2 M1 D1
Descripts

LO2 Discuss the implications of the use of data and information to support business processes in a
real-world scenario
Pass, Merit & Distinction P3 P4 M2 D1
Descripts

LO3 Explore the tools and technologies associated with data science and how it supports business
processes
Pass, Merit & Distinction P5 M3 D2
Descripts

LO4 Demonstrate the use of data science techniques to make recommendations to support real-
world business problems.
Pass, Merit & Distinction P6 P7 M4 D2
Descripts
Grade: Assessor Signature: Date:

Resubmission Feedback:

Grade: Assessor Signature: Date:

Internal Verifier’s Comments:

Signature & Date:

* Please note that grade decisions are provisional. They are only confirmed once internal and
external moderation has taken place and grades decisions have been agreed at the
assessment board.
Assignment Feedback
Formative Feedback: Assessor to Student

Action Plan

Summative feedback

Feedback: Student to Assessor

Assessor Date
signature

Student Date
signature
Pearson Higher Nationals in
Computing

Unit 17 - Business Process Support Assignment

General Guidelines

1. A Cover page or title page – You should always attach a title page to your assignment. Use
previous page as your cover sheet and make sure all the details are accurately filled.
2. Attach this brief as the first section of your assignment.
3. All the assignments should be prepared using a word processing software.
4. All the assignments should be printed on A4 sized papers. Use single side printing.
5. Allow 1” for top, bottom, right margins and 1.25” for the left margin of each page.

Word Processing Rules

1. The font size should be 12 point, and should be in the style of Time New Roman.
2. Use 1.5 line spacing. Left justify all paragraphs.
3. Ensure that all the headings are consistent in terms of the font size and font style.
4. Use footer function in the word processor to insert Your Name, Subject, Assignment No, and
Page Number on each page. This is useful if individual sheets become detached for any reason.
5. Use word processing application spell check and grammar check function to help editing your
assignment.

Important Points:
1. It is strictly prohibited to use textboxes to add texts in the assignments, except for the
compulsory information. eg: Figures, tables of comparison etc. Adding text boxes in the body
except for the before mentioned compulsory information will result in rejection of your work.
2. Carefully check the hand in date and the instructions given in the assignment. Late submissions
will not be accepted.
3. Ensure that you give yourself enough time to complete the assignment by the due date.
4. Excuses of any nature will not be accepted for failure to hand in the work on time.
5. You must take responsibility for managing your own time effectively.
6. If you are unable to hand in your assignment on time and have valid reasons such as illness, you
may apply (in writing) for an extension.
7. Failure to achieve at least PASS criteria will result in a REFERRAL grade.
8. Non-submission of work without valid reasons will lead to an automatic RE FERRAL. You will
then be asked to complete an alternative assignment.
9. If you use other people’s work or ideas in your assignment, reference them properly using
HARVARD referencing system to avoid plagiarism. You have to provide both in-text citation and
a reference list.
10. If you are proven to be guilty of plagiarism or any academic misconduct, your grade could be
reduced to A REFERRAL or at worst you could be expelled from the course
Student Declaration

I hereby, declare that I know what plagiarism entails, namely to use another’s work and to
present it as my own without attributing the sources in the correct form. I further understand
what it means to copy another’s work.

1. I know that plagiarism is a punishable offence because it constitutes theft.

2. I understand the plagiarism and copying policy of Pearson UK.
3. I know what the consequences will be if I plagiarise or copy another’s work in any of the
assignments for this program.
4. I declare therefore that all work presented by me for every aspect of my program, will
be my own, and where I have made use of another’s work, I will attribute the source in
the correct way.
5. I acknowledge that the attachment of this document signed or not, constitutes a binding
agreement between myself and Pearson UK.
6. I understand that my assignment will not be considered as submitted if this document is
not attached to the assignment.

Student’s Signature: Date:

(Provide E-mail ID) (Provide Submission Date)
Higher National Diploma in Business
Assignment Brief
Student Name /ID Number

Unit Number and Title Unit 17 - Business Process Support

Academic Year 2023/24

Unit Tutor

Assignment Title Business Process of a Company.

Issue Date

Submission Date

IV Name & Date

Submission format

The submission should be in the form of an individual report written in a concise, formal business
style using single spacing (refer to the assignment guidelines for more details). You are required to
make use of headings, paragraphs, and subsections as appropriate, and all work must be
supported with research and referenced using Harvard referencing system. Please provide in-text
citation and a list of references using Harvard referencing system.

The recommended word count is 4,500–5,000 words excluding annexures.

Minimum word count – 4,500
Maximum word count – 5,000
Unit Learning Outcomes:

LO1 Discuss the use of data and information to support business processes and the value they
have for an identified organization.
LO2 Discuss the implications of the use of data and information to support business processes in
a real-world scenario.
LO3 Explore the tools and technologies associated with data science and how it supports
business processes.
LO4 Demonstrate the use of data science techniques to make recommendations to support real-
world business problems.
Assignment Brief and Guidance:

Scenario

Select an organization of your choice and assume yourself as the newly recruited business
support executive to help the organization to enhance its business processes and decision-
making process using the latest data science tools and techniques.
Apply business support and data science tools and techniques into the context of the chosen
organization and complete the tasks given below.
The answers should be presented in a professionally compiled business report with appropriate
formatting and academic writing standards.

Task 1
Give a brief description to the organization chosen and its business processes.
Discuss how data and information support to run the business processes of the chosen
organization and the tools currently in use to manipulate meaningful data to support
organization’s business operations. Assess the value of data and information to the organization
and its individuals to run the business processes effectively and evaluate the implications of
using data and information to support the business processes. Your answer must include
examples where necessary from the chosen organization.

Task 2
Based on the nature of the chosen organization and its processes, analyse and discuss the
common threats, impacts, social, legal, ethical implications associated with data/information
use to support business processes. Describe how the threats and issues can be mitigated at a
personal and organizational level by proposing suitable solutions.
Task 3
Discuss how data science and the tools and technologies associated with it can be used support
business process and inform decisions of organizations by taking at least two examples from the
industry.
Identify a business problem or a requirement associate with decision making in the chosen
organization and evaluate how it could be addressed using data science tools and technologies.
Assess the benefits the organization and its users can receive through addressing the issue or
meeting the business requirement identified.

Task 4
Design and implement a data science solution to support decision making problem or the
requirement identified in task 3. Propose justified recommendations to improve the decision-
making process of the organization with the support of data science solution implemented.
Grading Rubric

Grading Criteria Achieved Feedback

LO1 Discuss the use of data and information to support business processes and the value they have for an identified organization

P1 Discuss how data and information support business

processes and the value they have for organizations.
P2 Discuss how data is generated and the tools used to
manipulate it to form meaningful data to support business
operations.
M1 Assess the value of data and information to
individuals and organizations in relation to real-world
business processes.
D1 Evaluate the wider implications of using data and
information to support business processes in an identified
organization.
LO2 Discuss the implications of the use of data and information to support business processes in a real-world scenario.

P3 Discuss the social legal and ethical implications of using

data and information to support business processes.
P4 Describe common threats to data and how they can be
mitigated at on a personal and organizational level.
M2 Analyze the impact of using data and information to
support business real-world business processes.
D1 Evaluate the wider implications of using data and
information to support business processes in an identified
organization.

LO3 Explore the tools and technologies associated with data science and how it supports business processes.
P5 Discuss how tools and technologies associated with
data science are used to support business processes and
inform decisions.

M3 Assess the benefits of using data science to solve

problems in real-world scenarios.

D2 Evaluate the use of data science techniques against

user and business requirements of an identified
organization.
LO4 Demonstrate the use of data science techniques to make recommendations to support real-world business problems.
P6 Design a data science solution to support decision
making related to a real-world problem.

P7 Implement a data science solution to support decision

making related to a real-world problem.

M4 Make justified recommendations that support

decision making related to a real-world problem.

D2 Evaluate the use of data science techniques against

user and business requirements of an identified
organization.
Table of Contents
1. Introduction
o 1.1 Purpose of the Report
o 1.2 Overview of Amazon
2. Task 1: Analysis of Amazon's Business Processes and Data Utilization
o 2.1 Description of Amazon's Business Processes
o 2.2 Role of Data and Information in Supporting Business Processes
o 2.3 Tools Currently in Use for Data Manipulation
3. Task 2: Analysis of Threats, Impacts, and Implications Associated with Data Use
o 3.1 Common Threats Related to Data Usage at Amazon
o 3.2 Social, Legal, and Ethical Implications of Data Use
o 3.3 Mitigation Strategies for Threats and Issues
 3.3.1 Personal Level Solutions
 3.3.2 Organizational Level Solutions
4. Task 3: Application of Data Science Tools and Techniques
o 4.1 Overview of Data Science Tools and Techniques
o 4.2 Industry Examples of Data Science in Business Process Enhancement
 4.2.1 Example 1: Predictive Analytics in Retail
 4.2.2 Example 2: Machine Learning for Inventory Management
o 4.3 Identification of a Business Problem at Amazon
 4.3.1 Problem: Improving Demand Forecasting
o 4.4 Evaluation of Data Science Tools to Address the Problem
o 4.5 Benefits of Implementing Data Science Solutions
5. Task 4: Design and Implementation of a Data Science Solution
o 5.1 Proposed Data Science Solution for Demand Forecasting at Amazon
 5.1.1 Data Collection and Preprocessing
 5.1.2 Model Selection and Implementation
6. Conclusion
7. References
Introduction

1.1 Purpose of the Report

This word aims to examine how Amazon, a world leader in e-commerce and technology, may apply cutting-edge data science tools and
methodologies to improve its business processes and decision-making capabilities. The purpose of this paper is to give a thorough overview of
the business procedures that Amazon now uses, the role that information and data play in supporting these procedures, and the instruments that
are currently in use for manipulating data.

1.2 Overview of Amazon

One of the biggest and most powerful corporations in the world, Amazon was founded by Jeff Bezos in 1994 and is best recognized for its e-
commerce platform. Nevertheless, Amazon's activities cover a broad spectrum of industries, including digital streaming (Amazon Prime
Video), cloud computing (Amazon Web Services), artificial intelligence (Alexa), and more. They also go far beyond online retail. With
millions of users depending on its services for anything from everyday shopping to cloud-based business solutions,
Amazon is a world

Amazon's business strategy is based on a dedication to operational efficiency, innovation, and client centricity. The business uses technology
and data to streamline its extensive supply chain, customize client interactions, and make data-driven decisions that strengthen its competitive
advantage. The capacity of Amazon to gather, examine, and act upon enormous volumes of data is central to its success, enabling it to
anticipate market trends, improve operational efficiency, and deliver exceptional customer service. (Amazon (Company), 2024)
 2. Task 1: Analysis of Amazon's Business Processes and Data Utilization

2.1 Description of Amazon's Business Processes

 The breadth and depth of Amazon's operations across multiple industries is reflected in the complexity and diversity of its business
procedures. The following major areas might be used to group the company's essential

2.1.1 Supply Chain Management

Amazon operates a vast and sophisticated supply chain network, encompassing procurement, warehousing, inventory management, and logistics.
The company sources products from suppliers worldwide, manages them in numerous fulfillment centers, and ensures timely delivery to
customers. The efficiency of this process is critical to maintaining Amazon’s competitive advantage, particularly in delivering products quickly
and cost-effectively.

2.1.2 Customer Relationship Management (CRM)

Amazon’s CRM processes are designed to optimize the customer experience by offering personalized services, handling customer inquiries, and
managing returns. The company utilizes a range of data-driven strategies to enhance customer satisfaction, such as personalized
recommendations, targeted marketing, and proactive customer service.

2.1.3 E-commerce Platform Operations

As the largest online retailer, Amazon’s e-commerce platform operations are central to its business. These processes include managing the
website, handling transactions, and ensuring a seamless user experience. The platform also supports third-party sellers, enabling them to list
products and manage sales through Amazon's marketplace.
 2.2 Role of Data and Information in Supporting Business Processes

 Data and information are the backbone of Amazon’s business processes. The company’s ability to collect, analyze, and utilize vast
amounts of data is crucial in optimizing operations, enhancing customer experiences, and driving innovation. Key roles of data and
information in Amazon’s processes include:

2.2.1 Enhancing Operational Efficiency

Amazon uses data to streamline its supply chain operations. Real-time data on inventory levels, customer orders, and supplier shipments enable
Amazon to minimize stockouts, reduce delivery times, and optimize warehouse management.

2.2.2 Personalizing Customer Experiences

Amazon’s recommendation engine is a prime example of how data supports business processes. By analyzing customer behavior, purchase
history, and browsing patterns, Amazon delivers personalized product recommendations, which significantly increase conversion rates and
customer loyalty.

2.2.3 Informed Decision-Making

Amazon leverages data to inform strategic decisions across all levels of the organization. For example, sales data, market trends, and customer
feedback are analyzed to guide decisions on product launches, pricing strategies, and marketing campaigns.

2.3 Tools Currently in Use for Data Manipulation

 Amazon employs a wide array of tools and technologies to manage and manipulate data effectively, ensuring that it can derive
actionable insights to support its business processes. Some of the key tools and platforms include:

2.3.1 Amazon Redshift

Amazon Redshift is a fully managed data warehouse service that allows Amazon to analyze vast amounts of data quickly and efficiently. It
supports complex queries and can scale to handle petabyte-scale data sets, making it ideal for business intelligence and analytics tasks.

2.3.2 Amazon Quick Sight

Amazon Quick Sight is a business intelligence tool that enables Amazon to create interactive dashboards and visualizations. This tool allows
users to explore data, generate reports, and gain insights without needing deep technical expertise.

2.3.3 Machine Learning and AI Tools

Amazon extensively uses machine learning (ML) and artificial intelligence (AI) tools, such as Amazon SageMaker, to build, train, and deploy
ML models. These models power various applications, from personalized recommendations to fraud detection and demand forecasting.
3. Task 2: Analysis of Threats, Impacts, and Implications Associated with Data Use

3.1 Typical Risks Associated with Amazon Data Usage

Being a data-driven company, Amazon must contend with a number of risks associated with data usage that could jeopardize business
operations, brand equity, and customer confidence. The most frequent dangers consist of:

3.1.1 Breach of Data

Similar to other sizable companies, Amazon is frequently the subject of cyberattacks, including data leaks. Sensitive consumer information,
including credit card numbers, purchase history, and personal information, may become public if unauthorized access is gained to Amazon's
enormous databases.

3.1.2 Violating Data Privacy

Given the volume of data that Amazon obtains from its users, there is always a chance that privacy laws will be broken. Penalties and damages
may arise from improper handling of data or noncompliance with data protection regulations, such as the General Data Protection Regulation
(GDPR).

3.2 Social, Legal, and Ethical Implications of Data Use

As a global leader in data-driven business operations, Amazon's use of data brings about significant social, legal, and ethical implications that
must be carefully managed to ensure the company's long-term success and maintain public trust.

3.2.1 Social Implications

The extensive collection and use of customer data by Amazon raise several social concerns:
  Privacy Concerns: Amazon’s ability to track customer behavior, preferences, and purchasing patterns can lead to feelings of intrusion
and surveillance among customers. The depth of insight Amazon has into individual lives may cause discomfort and fear of being
constantly monitored, potentially leading to a loss of trust in the company.
 Data-Driven Personalization and Bias: While data-driven personalization can enhance customer experiences, it can also lead to
unintended biases. For instance, recommendation algorithms may reinforce existing stereotypes or exclude certain demographic groups,
resulting in unequal access to products or services. This can have broader social consequences, such as perpetuating inequality or
marginalizing specific communities.
 Impact on Employment: The automation and data-driven decision-making processes at Amazon, particularly in logistics and
warehousing, have led to concerns about job displacement. As Amazon increasingly relies on robotics and AI, there is a social impact in
terms of reduced employment opportunities, particularly for low-skilled workers.

3.2.2 Legal Implications

Amazon must comply with a complex legal framework governing data usage across different jurisdictions:

• Data Protection Laws: Because Amazon operates in several nations, each has its own set of regulations pertaining to data protection, such as
the CCPA in California and the GDPR in Europe. Serious fines, legal action, and reputational harm to the business may arise from breaking
these requirements. According to these rules, Amazon must make sure that customers have control over their personal data and that it is securely
gathered, processed, and stored.

• Issues Regarding Intellectual Property: Amazon must manage intellectual property rights concerns as it develops new goods and services using
AI and massive databases. This includes ensuring that data used in machine learning models does not infringe on third-party copyrights or
patents, which could lead to costly legal disputes.

• Liability for Algorithmic Decisions: Amazon’s reliance on algorithms

3.3.1 Individual Level Resolutions

Amazon can take the following actions on an individual basis to enable staff members and clients to safeguard their information and reduce
risks:
Data Security Training: To increase understanding of data security best practices, all staff should get regular training. This entails being aware of
phishing efforts, managing sensitive data securely, and appreciating the value of data privacy.

Employee Access Controls: Strict access control procedures guarantee that only individuals with the proper authorization can access confidential
data. This lowers the possibility of employee misuse of illegal data and internal data breaches.

Tools for Customer Privacy: Amazon should give users the ability to manage their data, including privacy dashboards where they may evaluate
their data, change their preferences for data sharing, and refuse to receive targeted advertisements.

4.Task 3: Application of Data Science Tools and Techniques

4.1 Overview of Data Science Tools and Techniques

Large and complicated datasets can yield insightful information when used with a variety of tools, approaches, and techniques that are part of
the data science toolkit. Businesses such as Amazon are able to evaluate data, make well-informed choices, and streamline their operations
because to these technologies. The following are a few of the most widely used data science methods and tools:

 Predictive analytics is the process of forecasting future events based on past data. Regression analysis, time series forecasting, and
machine learning models are some of the methods frequently used to predict demand, trends, and consumer behavior.

 Data mining is the process of looking for patterns, correlations, and anomalies in big information. Methods such as association rule
mining, clustering, and classification can be used to find hidden information that influences business choices.

 The field of natural language processing, or NLP, is concerned with how computers and human language interact. NLP is used by
Amazon for speech recognition systems, chatbots, and sentiment analysis.

 Big Data Analytics: Big data analytics involves processing and analyzing massive datasets that traditional tools cannot handle. Tools
like Apache Hadoop, Spark, and NoSQL databases are essential for managing and analyzing big data.
  Visualization Tools: Data visualization tools like Tableau, Power BI, and Amazon Quick Sight help in creating interactive dashboards
and visual representations of data, making it easier for stakeholders to interpret and act on insights

4.2 Industry Examples of Data Science in Business Process Enhancement

Data science has revolutionized various industries by enhancing business processes and enabling more informed decision-making. Here are
two industry examples:

1. 4.2.1 An Example of Retail Predictive Analytics

 In the retail industry, predictive analytics is used to estimate customer demand, enhance inventory management, and optimize pricing
strategies. Retailers like Walmart and Target utilize predictive analytics as a method to forecast product demand by evaluating sales
data, consumer preferences, and outside variables like seasonality and economic conditions. This reduces unnecessary inventory, cuts
down on stockouts, and increases overall profitability.

 Anticipating which products are likely to be in high demand allows retailers to better plan their supply chain, allocate resources
efficiently, and employ dynamic pricing methods to increase sales and customer satisfaction.

2. 4.2.2 Second Example: Using Machine Learning to Manage Inventory

 Machine learning is used in manufacturing and logistics to enhance inventory control procedures. ML algorithms are used by UPS and
DHL, among other companies, to forecast inventory demands, streamline warehouse operations, and cut expenses. For instance, in
order to forecast future inventory needs, ML models can examine previous shipping data, order quantities, and supply chain
disturbances.

 This lowers the possibility of overstocking or stockouts, guarantees that stock levels match demand, and enables businesses to automate
replenishment procedures. Consequently, companies may raise customer satisfaction, lower holding costs, and increase operational
efficiency.
 4.3 Identification of a Business Problem at Amazon

4.3.1 Problem: Improving Demand Forecasting

Demand forecasting is one of Amazon's most important business issues. To satisfy consumer expectations, streamline supply chain
operations, and manage its massive inventory, Amazon needs accurate demand forecasts. The vastness and intricacy of Amazon's
operations mean that conventional forecasting techniques would find it difficult to keep up with the dynamic and quickly shifting market
circumstances.

4.4 Evaluation of Data Science Tools to Address the Problem

To address the problem of improving demand forecasting, Amazon can leverage several data science tools and techniques:

 Time Series Analysis: For forecasting based on historical data that takes trends and seasonality into consideration, tools like Prophet
(created by Facebook), SARIMA (Seasonal ARIMA), and ARIMA (Autoregressive Integrated Moving Average) work well.

 Machine Learning Models: To simulate intricate interactions between different elements influencing demand, Amazon can employ
machine learning techniques like Random Forest, Gradient Boosting, or Neural Networks. Even in the presence of non-linear
patterns and interactions, these models are able to learn from historical data and produce precise forecasts.

 Deep Learning: For sequential data forecasting, Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM)
networks can be very useful. This enables Amazon to include long-term patterns and dependencies into their demand forecasting
models.

4.5 Benefits of Implementing Data Science Solutions

 Implementing advanced data science solutions for demand forecasting can provide Amazon with several significant benefits:
  Improved Forecast Accuracy: By leveraging machine learning and time series analysis, Amazon can achieve more accurate demand
forecasts, reducing the likelihood of overstocking or stockouts. This leads to cost savings in inventory management and better resource
allocation.

 Enhanced Customer Satisfaction: Accurate demand forecasting ensures that products are available when customers need them, leading
to timely deliveries and a better overall shopping experience. This can result in higher customer satisfaction and increased loyalty.

 Optimized Supply Chain Operations: With better forecasts, Amazon can optimize its supply chain, from procurement to distribution,
ensuring that resources are used efficiently. This reduces operational costs and increases the speed and reliability of the supply chain.

5.1 Proposed Data Science Solution for Demand Forecasting at Amazon

5.1.1 Data Collection and Preprocessing

Data Collection

The first step in developing a demand forecasting model for Amazon is gathering relevant data. Given the vast amount of data Amazon handles,
the following data sources will be essential:

 Historical Sales Data: This includes past sales records across various product categories, customer segments, and geographic locations.
This data will be the primary input for demand forecasting models.
 External Data: Factors such as economic indicators, market trends, and competitor pricing can significantly influence customer demand.
Integrating external data sources like economic reports, social media trends, and market research can enhance the accuracy of the
forecasts.
 Customer Behavior Data: Analyzing customer browsing patterns, purchase history, and feedback can provide insights into changing
preferences and emerging trends, which are crucial for accurate demand forecasting.
 Inventory and Supply Chain Data: Data on current inventory levels, lead times, and supply chain performance will help ensure that the
forecasts are aligned with operational capabilities.
 Data Preprocessing

Once the data is collected, it must be preprocessed to ensure that it is clean, consistent, and ready for analysis. The preprocessing steps include:

 Data Cleaning: Handling missing values, outliers, and inconsistencies in the data is critical. Techniques such as interpolation for missing
values and outlier detection methods will be employed to ensure data quality.
 Feature Engineering: Creating new features that capture essential patterns in the data can significantly improve model performance. For
example, extracting features like seasonality indicators, promotional periods, and product lifecycle stages will help the model better
understand demand patterns.
 Data Normalization: Normalizing the data ensures that all features contribute equally to the model. Techniques like Min-Max scaling or
Z-score normalization will be applied to the data to standardize it.
 Splitting Data: The dataset will be split into training, validation, and test sets to evaluate the model's performance effectively. Typically,
an 80-10-10 split is used, where 80% of the data is used for training, 10% for validation, and 10% for testing.

5.1.2 Model Selection and Implementation

Model Selection

Given the complexity of demand forecasting at Amazon, selecting the right model is crucial. Several machine learning models will be evaluated,
including:

 Time Series Models:

o Prophet: Developed by Facebook, Prophet is a robust time series forecasting model that handles seasonality, holidays, and
missing data effectively. It is suitable for forecasting demand patterns over time.
o ARIMA/SARIMA: These traditional statistical models are effective for time series forecasting, especially when seasonality and
trends are well-defined.

 Machine Learning Models:

o Random Forest: A versatile model that can handle complex relationships and interactions in the data. Random Forest is known
for its robustness and ability to prevent overfitting.
 o Gradient Boosting Machines (GBM): GBM, including XGBoost and LightGBM, are powerful ensemble learning techniques that
can provide highly accurate predictions by combining the strengths of multiple decision trees.
o Neural Networks: Deep learning models like Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks
are particularly effective for sequential data and can capture long-term dependencies in demand patterns.

Implementation

Once the models are selected, they will be implemented and evaluated to determine which one performs best for Amazon's demand forecasting
needs. The implementation steps include:

 Model Training: The selected models will be trained on the historical sales data using the training dataset. Hyperparameter tuning will
be performed to optimize model performance.
 Model Validation: The validation dataset will be used to assess the model's ability to generalize to unseen data. Techniques like cross-
validation will be applied to ensure the model's robustness.
 Model Evaluation: The models will be evaluated using metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE),
and Mean Absolute Percentage Error (MAPE). The model with the lowest error metrics will be selected as the final model.
 Deployment: The best-performing model will be deployed into Amazon's production environment. This will involve integrating the
model with Amazon's existing systems, ensuring that it can process real-time data and provide timely demand forecasts.
 Continuous Monitoring and Improvement: The deployed model will be continuously monitored to ensure its accuracy and relevance.
Regular updates and retraining will be performed as new data becomes available, ensuring that the model adapts to changing market
conditions.

6. Conclusion
Amazon (company). (2024, August 15). Wikipedia. https://en.wikipedia.org/wiki/Amazon_(company )