Description Purpose
There are two parts in this assignment.
Part A provides you with opportunities to learn a range of machine learning methods and Python skills (GLO1 & ULO1) and apply your digital literacy to research and develop a machine learning solution (GLO3, GLO5, and ULO2). By completing this task, you will gain knowledge and skills in selecting and applying one or more appropriate machine learning algorithm(s) to develop and evaluate a machine learning solution and interpret the outcomes.
In Part B, you will report your application of machine learning and make recommendations to the business and management audience. By completing this task, you will gain ability to explain and justify machine learning options and discuss their pros and cons to the business audience.
Context/Scenario
The Great Ocean Banking Group, commonly known as Great Ocean Bank, a regional bank in Victoria, Australia, serves over 1 million customers and offers a comprehensive range of financial services predominantly to residents of Victoria and other states. With diverse offerings, including Home Mortgages, Personal Loans, and various savings and credit card accounts, the bank strives to meet the multifaceted financial needs of its customers. Through its periodic sales campaigns, it engages not only its existing clientele but also reaches out to prospective customers, highlighting the extensive benefits of its banking products and fostering a culture of financial well-being.
Aiming to enhance the effectiveness of these campaigns and to gain deeper insights into customer needs, the bank explores data analytics and Machine Learning opportunities to refine their marketing strategies and improve the overall customer experience, thereby reinforcing the Great Ocean Banking Group’s dedication to building lasting relationships and ensuring financial prosperity within the community.
You have been approached by Mr. Gary Peterson, Head of Customer Experience, and Ms. Kathy Hoang, Head of Data Analytics, at the Great Ocean Banking Group - the client. They are interested in exploring the dataset (GoBank.csv), which is a subset of data compiled from various databases of the Great Ocean Banking Group. This dataset includes customer information, banking relationships, details of the last contact from a marketing campaign, performances from a previous campaign, and several economic indices, such as the RBA cash rate. Further details are available in GoBank_metadata.csv.
Mr. Gary Peterson and Ms. Kathy Hoang seek to better understand their customers and the factors influencing their campaign performances, aiming to predict potential 'Sale' or 'No Sale' outcomes. You are tasked with performing two analytical tasks: uncovering data insights and exploring machine learning opportunities.
Regarding the former, Great Ocean Bank has requested an analysis on the following:
How do the demographic details of customers, such as Age and Qualification, among others, influence the success of sales outcomes (Sale Outcome)?
Does having any type of account (be it a home mortgage, personal loan, or any other account) influence a customer's decision to open an additional account with Great Ocean Bank during the current campaign, thereby impacting the Sale Outcome?
How does the method of our last contact with a customer (whether we called them - Outbound, or they called us - Inbound) influence their decision to open a new account (Sale) or not (No Sale)?
Does the outcome of previous campaigns influence the success of subsequent sales efforts (Sale Outcome)?
How do economic indicators (RBA Cash Rate, Employment Variation Rate, Consumer Confidence Index) impact Sale Outcomes?
Are there other things Great Ocean Bank hasn't considered yet that might affect whether their sales are successful, or do different factors influence each other?
Regarding the latter point—machine learning opportunities—the client is interested in exploring machine learning models. At this stage, they require one predictive machine learning model to predict potential 'Sale' or 'No Sale' outcomes and one clustering model to be developed. Suggestions for future opportunities are also welcome.
Datasets provided:
GoBank.csv
Data description
GoBank_metadata.csv
You are required to explore the first dataset, GoBank.csv, and develop and test machine learning option(s) using Python. You are also required to develop two reports:
The first technical report (Part A) should present your analysis and findings to Kathy Hoang, Head of Data Analytics. This report should detail your approach to exploring the dataset, the machine learning techniques used, and your findings. Your findings should be supported by relevant visualisations and statistical analysis. This report should also develop and compare two predictive machine learning models using different algorithms, recommend one of these models; develop and evaluate one clustering model; inform model deployment, and recommend future engagements with the client. See further details in the Specific Requirements section below.
Based on your data analysis and machine learning findings, you are then required to develop a consultancy report (Part B) for Gary Peterson, Head of Customer Experience. The report should include your response to the client's six (6) questions, the proposed machine learning models, and recommendations for use. You should also discuss the limitations of your approach and any potential areas for future improvements. See further details in the Specific Requirements section below.
See further details in the section Specific Requirements below.
The dataset, acquired from a public source, has been modified and pre-processed for the assessment purpose.
Specific Requirements
You are required to:
Develop your business and data understandings.
Prepare and explore the provided dataset, cleanse and pre-process data as needed. Undertake an exploratory data analysis (EDA) to respond to the client’s six questions.
Undertake supervised machine learning model development, evaluation, and selection. Two predictive models should be developed, tested, and compared.
Undertake unsupervised machine learning using clustering analytics.
Develop two reports:
- The first technical report (Part A) should present your EDA (Exploratory Data Analysis) and machine learning findings to Kathy Hoang, Head of Data Analytics.
- The second consultancy report (Part B), for Gary Peterson, Head of Customer Experience, should present, in the business domain language, responses to the six specific requests about data, insights from clustering analytics, and one predictive machine learning model.
Format and present your report professionally. Two sample report templates are provided under Assessment Resources.
Correctly use the APA7 style of referencing, and include in-text citations when quoting, referring to, summarising, or paraphrasing from any source.
Report Sample PART A
CONTEXT
The Great Ocean Banking Group, commonly known as Great Ocean Bank, a regional bank in Victoria, Australia, serves over 1million customers and offers a comprehensive range of financial services predominantly to residents of Victoria and other states. With diverse offerings, including Home Mortgages, Personal Loans, and various savings and credit card accounts, the bank strives to meet the multifaceted financial needs of its customers. Through its periodic sales campaigns, it engages not only its existing clientele but also reaches out to prospective customers, highlighting the extensive benefits of its banking products and fostering a culture of financial well-being.
Aiming to enhance the effectiveness of these campaigns and to gain deeper insights into customer needs, the bank explores data analytics and Machine Learning opportunities to refine their marketing strategies and improve the overall customer experience, thereby reinforcing the Great Ocean Banking Group’s dedication to building lasting relationships and ensuring financial prosperity within the community.
This assignment is divided into two parts aimed at enhancing machine learning skills and applying them to solve a business problem for Great Ocean Bank. In Part A, the focus is on learning various machine learning methods and Python programming skills to develop a machine learning solution. This involves selecting suitable algorithms, applying them to the dataset, evaluating their performance, and interpreting the outcomes. Part B involves reporting the application of these machine learning techniques to a business audience, making recommendations based on the analysis, and discussing the pros and cons of the chosen methods. Great Ocean Bank, a regional bank in Victoria, Australia, serves over 1 million customers and offers a variety of financial services, including home mortgages, personal loans, savings accounts, and credit cards. The bank conducts periodic sales campaigns to engage with both existing and potential customers, aiming to enhance customer experience and foster financial well-being. To improve the effectiveness of these campaigns, the bank is exploring data analytics and machine learning opportunities. Mr. Gary Peterson, Head of Customer Experience, and Ms. Kathy Hoang, Head of Data Analytics, have tasked us with analyzing a dataset from their databases to better understand the factors influencing campaign performance and to predict 'Sale' or 'No Sale' outcomes
Great Ocean Bank, a prominent regional bank in Victoria, Australia, is dedicated to serving over 1 million customers with a broad array of financial services. These services include home mortgages, personal loans, savings accounts, and credit card accounts, catering to the diverse financial needs of both residents of Victoria and other states. The bank conducts periodic sales campaigns, aiming to engage not only its existing clientele but also prospective customers. These campaigns highlight the extensive benefits of Great Ocean Bank's products and services, fostering a culture of financial well-being within the community. In the competitive banking industry, understanding customer behavior and preferences is crucial for the success of marketing campaigns. The bank's management, represented by Mr. Gary Peterson, Head of Customer Experience, and Ms. Kathy Hoang, Head of Data Analytics, seeks to enhance the effectiveness of these campaigns. They aim to gain deeper insights into customer needs and the factors influencing their decisions to open new accounts. This understanding will enable the bank to refine its marketing strategies, improve customer engagement, and ultimately increase the success rate of its sales campaigns.
The dataset provided by Great Ocean Bank includes comprehensive customer information, details of banking relationships, last contact details from marketing campaigns, outcomes of previous campaigns, and several economic indicators such as the RBA cash rate. Analyzing this dataset provides an opportunity to uncover patterns and insights that can inform the bank's marketing efforts.
The primary business objectives are to:
Enhance Customer Understanding: By analyzing demographic details, previous campaign outcomes, and economic indicators, the bank aims to better understand the factors that drive customer decisions. This deeper understanding will help tailor marketing strategies to individual customer profiles, improving the relevance and effectiveness of the campaigns.
Improve Campaign Effectiveness: Identifying the most influential factors that lead to successful sales outcomes will allow the bank to focus its efforts on high-potential segments. By understanding which customers are more likely to respond positively, the bank can allocate resources more efficiently and improve the overall success rate of its campaigns.
Optimize Contact Strategies: Analyzing the impact of contact methods (outbound vs. inbound) on sales outcomes will help the bank develop more effective communication strategies. This insight can guide the timing, frequency, and method of customer interactions, enhancing customer experience and increasing the likelihood of successful sales.
Leverage Economic Indicators: Understanding the influence of broader economic conditions on customer behavior will enable the bank to adjust its strategies in response to changes in the economic environment. This proactive approach can help mitigate risks and capitalize on opportunities presented by economic trends.
The dataset provided by Great Ocean Bank consists of 22,940 entries and 18 columns, containing various features related to customer demographics, banking relationships, and campaign details. The primary objective is to predict the Sale Outcome, indicating whether a customer opened a new account following a sales campaign. Dataset has 22,940 Number of Rows & 18 Number of Columns.
Feature Information:
Age: Customer's age (numerical, no missing values).
Qualification: Customer's educational qualification (categorical, 162 missing values).
Occupation: Customer's occupation (categorical, no missing values).
Marital Status: Customer's marital status (categorical, no missing values).
Home Mortgage: Whether the customer has a home mortgage (categorical, no missing values).
Personal Loan: Whether the customer has a personal loan (categorical, no missing values).
Has Other Bank Account: Whether the customer has another bank account (categorical, no missing values).
Last Contact Direction: The direction of the last contact (inbound or outbound) (categorical, 47 missing values).
Last Contact Duration: Duration of the last contact in seconds (numerical, 141 missing values).
Last Contact Month: Month of the last contact (categorical, no missing values).
Last Contact Weekday: Weekday of the last contact (categorical, no missing values).
Number of Current Campaign Calls: Number of calls made during the current campaign (numerical, no missing values).
Number of Previous Campaign Calls: Number of calls made during previous campaigns (numerical, 26 missing values).
Previous Campaign Outcome: Outcome of the previous campaign (categorical, 206 missing values).
RBA Cash Rate: The Reserve Bank of Australia's cash rate (numerical, no missing values).
Employment Variation Rate: Employment variation rate (numerical, no missing values).
Consumer Confidence Index: Consumer confidence index (numerical, no missing values).
Sale Outcome: Target variable indicating the outcome of the sales process (categorical, no missing values).
Target Variable: The target variable, Sale Outcome, represents whether a customer has opened a new account (Sale) or not (No Sale) following a sales campaign. It is a categorical variable used to evaluate the success of the bank's marketing strategies.
This dataset provides a comprehensive overview of the factors influencing customer behavior and the success of sales campaigns at Great Ocean Bank. By analyzing these features, we aim to develop a machine learning model to predict sales outcomes and offer actionable insights to improve future campaigns.
Feature Types
- Categorical Features: 11 columns including qualification, occupation, and previous campaign outcomes.
- Numerical Features: 7 columns including age, contact duration, and economic indicators.
Null Values
- Qualification: 162 null values.
- Last Contact Direction: 47 null values.
- Last Contact Duration: 141 null values.
- Number of Previous Campaign Calls: 26 null values.
- Previous Campaign Outcome: 206 null values.
1. Age Distribution:
- The mean age of the individuals in the dataset is approximately 40 years.
- The minimum age is 17 years, and the maximum age is 98 years.
- Most individuals fall between the age range of 32 to 47 years, as indicated by the 25th and 75th percentiles.
2. Last Contact Duration:
- The mean duration of the last contact is around 282 seconds.
- The shortest contact duration recorded is 0 seconds, while the longest is 4918 seconds.
- There's a considerable variation in contact durations, with a standard deviation of approximately 286 seconds.
3. Campaign Calls:
- On average, individuals received about 2.5 campaign calls during the current campaign.
- The number of previous campaign calls has a mean of approximately 0.2, indicating a low average number of calls from previous campaigns.
4. Economic Indicators:
- The dataset includes economic indicators such as RBA Cash Rate, Employment Variation Rate, and Consumer Confidence Index.
- These indicators show relatively consistent values across the dataset, with varying degrees of deviation from the mean.
Data Preparation steps involved both removing unnecessary columns and handling missing values effectively, The missing values are filled with the median value of the column to ensure completeness of the dataset for analysis.
1. Removal of Unwanted Columns:
- One extra columns, namely CustomerID were identified as irrelevant to the analysis and were subsequently removed from the dataset using the `drop()` function.
2. Handling Missing Values:
- We have some missing values which we have removed directly as of no use we can directly remove NA values no such huge impact on data
By removing unnecessary columns and addressing missing values, the dataset was effectively prepared for further exploration and analysis. These data preparation steps ensure the dataset's integrity and reliability, enabling more accurate insights and predictions to be derived from the data.
- The histogram represents the distribution of ages in a dataset. Most individuals fall within the age range of 25 to 64. The distribution is unimodal, with a peak around age 50.
1. Qualification:
- Most people have a secondary education, followed by tertiary education. Primary education is the least common
- Consider exploring the reasons behind this distribution—such as regional factors or educational policies.
2. Occupation:
- The majority of people are blue-collar workers, while management positions also have a significant representation.
- Entrepreneurs and housemaids are among the occupations with the fewest individuals.
3. Marital Status:
- A large number of individuals are married, with single individuals also making up a significant portion.
- Divorced individuals are fewer in comparison.
- Explore potential correlations between marital status and other variables.
4. Home Mortgage:
- The data is divided into three categories, but it's not clear what these categories represent due to the lack of labels.
- Consider labeling the categories (e.g., "No Mortgage," "Partial Mortgage," "Full Mortgage") for better interpretation.
5. Personal Loan:
- Similar to the home mortgage graph, there's no clear indication of what "yes" or "no" refers to.
- However, there’s a higher count for “no” than “yes.”
6. Has Other Bank Account:
- A significant number of people do not have another bank account compared to those who do.
- Investigate factors influencing this distribution (e.g., accessibility, financial literacy).
7. Last Contact Direction:
- More people were contacted outbound (by the bank) than inbound (by the customer).
- Consider analyzing the effectiveness of each contact direction.
8. Last Contact Month:
- May has the highest contact rate, while December has the lowest.
- Explore seasonality effects or campaign strategies during specific months.
9. Last Contact Weekday:
- Contacts seem fairly evenly distributed across weekdays, with Monday having slightly more contacts.
- Investigate whether certain weekdays yield better response rates.
10. Previous Campaign Outcome:
- The majority were not contacted in a previous campaign.
- Understand the impact of previous campaign outcomes on the current campaign strategy.
1. Qualification:
- People with professional qualifications Bachelor degree have higher sales outcomes, followed by those with secondary education.
- Consider exploring regional or industry-specific factors that contribute to this trend.
2. Occupation:
- Sales are more common among management professionals and technicians.
- Investigate why certain occupations exhibit higher sales rates.
3. Marital Status:
- Married individuals have a higher rate of sales compared to single or divorced individuals.
- Explore whether marital status impacts purchasing decisions.
4. Has Other Bank Account:
- People without other bank accounts have made more purchases.
- Explore reasons behind this trend (e.g., financial stability).
5. Last Contact Direction:
- Outbound contacts (by the bank) result in more successful sales than inbound ones.
- Consider optimizing contact strategies.
6. Last Contact Month:
- March shows the highest number of successful sales, while December has the least.
- Explore seasonality effects or campaign timing.
We are providing your task solution with complete code explanation. If you need complete solution of above problem statement or need other research project help then you can contact us. We are providing high quality work as per your given instructions.
For more details you can send your project requirement details at:
Comments