TECH3200 Artificial Intelligence and Machine Learning in IT Report 3 Sample
Assignment Details
Your Task
Your third assessment requires you to analyse the supplied business case and understand the problem of the business so that you, as the subject matter expert in Machine Learning will write a solution proposal to highlight how your solution is suitable for supporting the business in several areas.
Description
Proposal is a formal written offer from a seller to a prospective sponsor. In this assessment, it is a solution proposal that you will write to advocate your ML solution and to convince the Board of the business that your solution is the best. Apart from understanding the business requirements in the business case, you will also need to conduct research on AI/ML including their benefits to the industry where the business is in, how AI/ML is going to support the specified areas highlighted in the business case. You need to have a good understanding of ML algorithms and applications used in modelling e.g., prediction.
Business Case
YourSuperStore is a market leader in grocery retail. It has over 2000 stores nationwide covering most of the major cities. It is becoming more and more important to fulfill people’s daily grocery needs. Due to the significant growth of the business. The old Loyalty Program that has over 3 million members, is no longer suitable to support the business. The board has decided to implement a new Loyalty Program (hint: it is like “Everyday Rewards” in Woolworths) to advance the business in various areas:
• Customer retention
• Marketing communications
• Real-time offers
• Personalised rewards and pricing
• Predictive analytics (cross-selling and up-selling)
• Fraud detection
AI and Machine Learning are rapidly transforming many industries including grocery retail. YourSuperStore’s competitors have quickly made the move to invest in trending AI technologies. To retain its market-leading position, the business would like to incorporate machine learning into the new Loyalty Program.
The CEO has appointed you as the subject matter expert in Machine learning knowing the algorithms and applications to work with the project team for the Loyalty Program.
Proposal
Before project funding can be approved, you are required to write a solution proposal to outline:
• Executive summary
• Objectives
• Business problems
• Benefits of AI/ML
• ML solution for the Program: (use ML algorithms, applications for data manipulation and modelling to justify how your solution will support the following business areas respectively)
o Customer retention
o Marketing communications
o Real-time offers
o Personalised rewards and pricing
o Predictive analytics (cross-selling and up-selling)
o Fraud detection
• Challenges
• Potential ethical and security issues
• Recommendations
Referencing
Add your references (at least 5) on the last page using any professional and consistent styling.
Solution
Executive Summary
Machine Learning and Artificial Intelligence have nowadays become the most trending topics or aspects in almost every place, especially in business. Operational head of YourSuperStore, which is a famous grocery retailer, intends of include machine learning into the new Loyalty Program of the organization. Therefore, this paper is all about preparing a proposal depicting algorithms and their applications to work with the project team for the loyalty program of YourSuperStore. Alongside exploring the benefits of AI and ML, this study has also proposed ML solution for the program by using ML algorithms, modelling to justify how the solution would support several business areas like fraud detection, customer retention, real-time offers and so on.
Objectives
YourSuperStore lays special stress on retaining its leading position in the grocery retain market. Thus, the proposal has the following objectives.
1. To recognize the problems faced by YourSuperStore and to explore the significance of AI or ML to solve the problems
2. To incorporate ML solution for the new Loyalty Program
3. To explore challenges, ethical and security constraints along with establishing some recommendations.
Business Problems
YourSuperStore is one of the most famous grocery retailers in the world. However, the previous Loyalty Program of this organization has become obsolete, and it is not capable of supporting the business operations of this company. It has become very difficult to earn rewards. Sometimes, it takes a lot of time to earn rewards from the customer end. Therefore, this organization has decided to emphasize developing new Program where some areas like fraud detection, predictive analytics, personalized rewards and pricing, real-time offers, marketing communications and customer retention. With passing time, it is becoming very crucial to reevaluate the Loyalty Program of YourSuperStore and modernize it by including new features or updating the logic of loyalty for university assignment help.
Benefits of AI or ML
Both AI and ML effectively participate in transforming YourSuperStore’s Loyalty Program by enabling personalized rewards in accordance with the preferences of customers. YourSuperStore can also use ML or AI algorithms for generating predictive insights, which can help companies for making better informed decisions. This organization can recognize patterns and trends in consumer behavior by helping them in forecasting future sales and making data-driven decisions (Baidoo-Anu & Ansah, 2023, p. 54). On the other hand, these algorithms can also be utilized for analyzing huge amounts of data, correlations and patterns, which would help businesses in making decisions. AI and ML helps in inducing real-time offers that can improvise the loyalty programs by giving personalized incentives and rewards according to the consumer behavior.
ML Solutions
Customer Retention
Churn prediction is a significant application of Machine Learning for customer retention. This process recognizes which consumers intend to stop the use of the products of YourSuperStore in the near future (Lalwani et al., 2022, p. 22). This organization can implement churn prediction model by passing previous data of customers through a model machine learning for recognizing the connections between targets and features and can make predictions regarding the new customers and their behaviors (Ullah et al., 2019, p. 60135). Apart from that, propensity data modelling technique participates in predicting the performance of consumers by analyzing their previous behavioral patterns (Adelson, 2019, p. 15). YourSuperStore can develop robust propensity models along with making accurate forecasts by applying the concepts of Machine Learning. The propensity score is generally estimated with the use of logistic regression model. Logistic regression can be proposed to be used as the ML algorithm for YourSuperStore. This model helps in estimating the probability of the occurrence of an event depending on the dataset of independent variables. The range of dependent variables varies between 0 to 1 (Gasso, 2019, p. 25). YourSuperStore can consider customer characteristics like age, profession and shopping habits of customers and marketing activities like advertising and sales events as the independent variables whereas retention actions and churn are taken into account as the dependent variables.
Marketing Communication
When improvement of marketing communication is concerned, network modelling technique comes into play an important role. It works on the concepts of graph theory along with its mathematical contexts that investigates the relationship between nodes and edges (Srinidhi, Ciga & Martel, 2021, p. 101813). YourSuperStore would definitely find it very useful to create mathematical presentations of data marketing to uncover patterns, relations and trends.
YourSuperStore should implement K-means clustering algorithm because it would help this organization for extracting actionable information from the data of customers along with making decisions driven by data, which drive success and growth. This algorithm clusters data in several groups and provides convenience to discover group categorization in dataset unlabeled on its own without the necessity of any training (Ismkhan, 2018, p. 410).
.png)
Fig 1: K-Means Clustering Algorithm
(Source: Ahmed, Seraj & Islam, 2020, p. 1295)
Real-time offers
YourSuperStores should apply real-time processing in order to process data at a near-instant rate needing a continuous data flow including both the output and intake data for maintaining real-time insights. This method would be effective for this organization because real-time data processing has the capacity of enhancing operations, optimizing business outcomes, visibility for IT architecture, boosting monitoring and even improving the overall customer experiences (Habeeb et al., 2019, p. 300).
YourSuperStores can apply the Naïve Bayes Algorithm as it is a famous supervised machine learning algorithm utilized for the tasks of classification like text classification. The application of this algorithm would be useful for this organization as it efficiently participates in categorizing new observations into the predefined classes for the limitless data (Saritas & Yasar, 2019, p. 90). In Machine Learning, it is the fastest model that has immense efficiency in making fast predictions.
.png)
Fig 2: Naïve Bayes Algorithm
(Source: Rahat, Kahir & Masum, 2019, p. 268)
Personalized Rewards and Pricing
YourSuperStore should go for adopting a dimensional model in the aspect of personalized rewards and pricing. It works on depicting business processes throughout a company and organizes that data and its structure in a logical way (Ettl et al., 2020, p. 462). It would conduct reporting, analysis, and query operations for this organization. YourSuperStore should improvise personalized rewards and pricing aspects by applying decision tree algorithm that takes part in determining the alternative that would yield the biggest expected gain in the monetary context (Patel & Prajapati, 2018, p. 76). In addition to this, the algorithm also assures that the information and alternatives are helpful for business decision-making.
Predictive analytics
As YourSuperStore operates in the grocery retail industry, this organization should lay special stress on performing predictive analytics in cross-selling by giving views into the behavior of customers. It would be easier for the organization to predict which customers are most likely to respond to a specific service or product (Bertsimas & Kallus, 2020, p. 1030). A predictive data modelling process would be useful for the company to retain its business performance within a highly competitive marketplace. Moreover, clustering algorithm should be used by YourSuperStore as it deals with recognizing distinct customer segments (Rodriguez et al., 2019, p. 210236). By applying this algorithm, YourSuperStore can effectively tailor its customer support, product offerings and marketing strategies for better meeting the needs of every group. This algorithm can automatically recognize the pattern inside the data to analyze the collected data without their labels.
Fraud Detection
YourSuperStore should apply the random forest algorithm as it acts as the binary classifier that can make it suitable for the detection of frauds like credit card frauds. Lin and Jiang, (2021, p. 2683) have stated that random forest is such a versatile and potential machine learning technique that combines and grows several decision-trees for creating a forest. This method is taken into account as a supervised learning algorithm utilized for regression and classification tasks.
In case of detecting frauds, predictive modelling method can be useful for YourSuperStore because it efficiently detects potential frauds by analyzing past patterns and data for identifying suspicious activities (Carcillo et al., 2021, p. 330). YourSuperStore can also anticipate its future outcomes along with taking proactive measures for the prevention and detection of frauds.
Challenges
Application of Machine Learning is not above problems, which are as follows:
1. Massive data sets are required by ML for the purpose of training and they should be unbiased or inclusive. Moreover, there can also be some instances where they would have to wait for new data to be generated.
2. Machine Learning also requires sufficient time to help the algorithms in developing and learning the proper way of fulfilling their purposes on the basis of relevancy and accuracy. It also requires huge resources for functioning in an appropriate way (Karpatne et al., 2018, p. 1550). It can mean extra necessities of computer power for an organization or an operator.
3. ML is autonomous but it is susceptible to errors. If an algorithm is trained with data sets that are small enough to not be inclusive. Hence, biased predictions would come up from a biased training set (D'Amour et al., 2022, p. 10237). It can result in irrelevance in advertisements being displayed to customers. Such blunders in the case of ML can set off several errors that can remain unidentified for quite a long span of time.
Potential Ethical and Security Issues
1. Updates of ML systems raise the level of efficiency of these approaches in such a way that they would need minimal human control. A great risk to the process of decision-making procedure considered by a business organization can be created by it (Karpatne et al., 2018, p. 1550).
2. Gap in interpretability is taken into account as the black box issue. It can generally be seen in more complex processes of ML such as neutral networks.
3. Machine Learning can also invoke some security attacks by enabling powerful impersonation of social engineering and phishing attacks.
4. Security issues like adversarial inputs, adversarial inference, poisoning of training data and model inversion can significantly expose the training data or model parameters (D'Amour et al., 2022, p. 10237).
5. A significant threat is also posed by the algorithmic bias to justice and fairness in the processes of decision making that heavily rely on the systems of Artificial Intelligence. Biased algorithms utilized in the processes of recruitment may unfairly favor some specific candidates while discriminating against others depending on factors unrelated to their qualifications.
Recommendations
To ensure successful deployment of Machine Learning system, YourSuperStore should incorporate some important recommendations, which are as follows:
1. This organization should be more conscious about the control and encryption of the access to the data of training by applying continuous update on the systems for new threats, implementation of stricter controls, use of the adversarial training for hardening of models and so on.
2. Machine Learning’s overfitting should be mitigated using efficient training processes or increasing the samples’ diversity. Moreover, YourSuperStore should address the underfitting problem by incorporating more input features.
3. YourSuperStore should improve the performance of the ML system by applying Grid search. It focuses on identifying the optimal set of hyperparameters by the evaluation of all the combinations.
Reference List
Adelson, J.L., 2019. ‘Educational research with real-world data: Reducing selection bias with propensity score analysis’ Practical Assessment, Research, and Evaluation, vol. 18, no. 1, p.15.
Ahmed, M., Seraj, R. & Islam, S.M.S., 2020. ‘The k-means algorithm: A comprehensive survey and performance evaluation’. Electronics, vol. 9, no. 8, p.1295.
Baidoo-Anu, D. & Ansah, L.O., 2023. ‘Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning.’, Journal of AI, vol. 7, no. 1, pp.52-62.
Bertsimas, D. & Kallus, N., 2020. ‘From predictive to prescriptive analytics’. Management Science, vol. 66, no. 3, pp.1025-1044.
Carcillo, F., Le Borgne, Y.A., Caelen, O., Kessaci, Y., Oblé, F. & Bontempi, G., 2021. ‘Combining unsupervised and supervised learning in credit card fraud detection’. Information sciences, vol. 557, pp.317-331.
D'Amour, A., Heller, K., Moldovan, D., Adlam, B., Alipanahi, B., Beutel, A., Chen, C., Deaton, J., Eisenstein, J., Hoffman, M.D. & Hormozdiari, F., 2022. ‘Underspecification presents challenges for credibility in modern machine learning’. The Journal of Machine Learning Research, vol. 23, no. 1, pp.10237-10297.
Ettl, M., Harsha, P., Papush, A. & Perakis, G., 2020. ‘A data-driven approach to personalized bundle pricing and recommendation’. Manufacturing & Service Operations Management, vol. 22, no. 3, pp.461-480.
Gasso, G., 2019. ‘Logistic regression’. INSA Rouen-ASI Departement Laboratory: Saint-Etienne-du-Rouvray, France, pp.1-30.
Habeeb, R.A.A., Nasaruddin, F., Gani, A., Hashem, I.A.T., Ahmed, E. & Imran, M., 2019. ‘Real-time big data processing for anomaly detection: A survey’. International Journal of Information Management, 45, pp.289-307.
Ismkhan, H., 2018. ‘Ik-means−+: An iterative clustering algorithm based on an enhanced version of the k-means’. Pattern Recognition, vol. 79, pp.402-413.
Karpatne, A., Ebert-Uphoff, I., Ravela, S., Babaie, H.A. & Kumar, V., 2018. ‘Machine learning for the geosciences: Challenges and opportunities’. IEEE Transactions on Knowledge and Data Engineering, vol. 31, no. 8, pp.1544-1554.
Lalwani, P., Mishra, M.K., Chadha, J.S. & Sethi, P., 2022. ‘Customer churn prediction system: a machine learning approach’. Computing, pp.1-24.
Lin, T.H. & Jiang, J.R., 2021. ‘Credit card fraud detection with autoencoder and probabilistic random forest’. Mathematics, vol. 9, no. 21, p.2683.
Patel, H.H. & Prajapati, P., 2018. ‘Study and analysis of decision tree-based classification algorithms’, International Journal of Computer Sciences and Engineering, vol. 6, no. 10, pp.74-78.
Rahat, A.M., Kahir, A. & Masum, A.K.M., 2019. ‘Comparison of Naive Bayes and SVM Algorithm based on sentiment analysis using review dataset’. In 2019 8th International Conference System Modeling and Advancement in Research Trends (SMART), pp. 266-270.
Rodriguez, M.Z., Comin, C.H., Casanova, D., Bruno, O.M., Amancio, D.R., Costa, L.D.F. & Rodrigues, F.A., 2019. Clustering algorithms: A comparative approach. PloS one, vol. 14, no. 1, p.e0210236.
Saritas, M.M. & Yasar, A., 2019. ‘Performance analysis of ANN and Naive Bayes classification algorithm for data classification’. International journal of intelligent systems and applications in engineering, vol. 7, no. 2, pp.88-91.
Srinidhi, C.L., Ciga, O. & Martel, A.L., 2021. Deep neural network models for computational histopathology: A survey. Medical Image Analysis, 67, p.101813.
Ullah, I., Raza, B., Malik, A.K., Imran, M., Islam, S.U. & Kim, S.W., 2019. ‘A churn prediction model using random forest: analysis of machine learning techniques for churn prediction and factor identification in telecom sector’. IEEE access, vol. 7, pp.60134-60149.
Would you like to schedule a callback?
Send us a message and we will get back to you
Highlights
Earn While You Learn With Us
Confidentiality Agreement
Money Back Guarantee
Live Expert Sessions
550+ Ph.D Experts
21 Step Quality Check
100% Quality
24*7 Live Help
On Time Delivery
Plagiarism-Free
81 Isla Avenue Glenroy, Mel, VIC, 3046 AU
