opcode-open-spring-fest/hidden-consumer-patterns

The Hidden Language of Consumers: Decoding Unseen Patterns

hidden-consumer-patterns's Issues

[Feature]: Appropriate EnrollmentDate Format

So, what is it about?

The current format of the EnrollmentDate feature is object, leading to errors that may affect data processing and analysis.

Objective: To convert the EnrollmentDate field into a consistent and standardized date format that aligns with our data handling practices and system requirements.

Code of Conduct

• I agree to follow this project's Code of Conduct

[Feature]: Bivariate Analysis of Engineered Features with Parental Status Hue

So, what is it about?

Overview

This issue proposes the implementation of a bivariate analysis of engineered features within our customer dataset, utilizing is_parent as a hue to distinguish between parental statuses. The goal is to combine insights from this analysis with previous univariate analyses to extract key findings that can drive informed business decisions.

Objectives

• To perform a feature correlation analysis to understand the relationships between different customer attributes.
• To conduct a bivariate analysis with a focus on the 'is_parent' feature to observe how parental status may affect other variables.
• To synthesize the results from this bivariate analysis with prior univariate analysis to compile a comprehensive report on customer behavior and trends.

Expected Outcomes

• A correlation matrix that highlights significant correlations between features.
• Visual representations (scatter plots, pair plots) that showcase the bivariate relationships with 'is_parent' as a hue.
• A summary of key insights that have been derived from the combination of univariate and bivariate analyses, potentially revealing patterns unique to parents or non-parents within the dataset.

Code of Conduct

• I agree to follow this project's Code of Conduct

[Feature]: AnnualIncome Analysis by AcademicLevel and Statistical Testing

So, what is it about?

This issue focuses on the need for a detailed analysis of AnnualIncome by AcademicLevel. The task involves the following steps:

1. AnnualIncome by AcademicLevel Plot: Generate a plot to visualize the distribution of AnnualIncome across different AcademicLevels. This will help us understand if there are significant differences in income levels across different academic qualifications.

2. Statistical Test: Perform a statistical test to check if the average AnnualIncome is the same for PHD owners and Master degree owners.

3. Null and Alternate Hypothesis Statements: Clearly state the null and alternate hypotheses for the statistical test. The null hypothesis could be that the average AnnualIncome is the same for both PHD and Master degree owners, while the alternate hypothesis could be that the average AnnualIncome is not the same for these two groups.

4. Significance Level and Test Method: Specify the significance level chosen for the test (for example, 0.05) and detail the test method used.

This issue is crucial for understanding the impact of academic qualifications on income levels. Any progress or updates regarding this issue will be posted here.

Code of Conduct

• I agree to follow this project's Code of Conduct

[Feature]: Univariate Plots, Outlier Detection and Removal in Engineered Features

So, what is it about?

This issue focuses on the need for visualizing the engineered features and AnnualIncome using univariate plots. The task involves the following steps:

1. Univariate Plots: Generate univariate plots for the engineered features to understand their individual characteristics.

2. Outlier Detection: Implement outlier detection using boxplots. This will help identify any extreme values that may adversely affect our analysis.

3. Dealing with Outliers: Once outliers are detected, devise a strategy to handle them. This could involve removing these outliers or adjusting them to fall within an acceptable range.

4. Final Plot without Outliers: After dealing with the outliers, generate the final univariate plots without the outliers. This will give us a more accurate representation of the engineered features.

This issue is crucial for ensuring the quality and reliability of our data analysis. Any progress or updates regarding this issue will be posted here.

Code of Conduct

• I agree to follow this project's Code of Conduct

[Feature]: Deal with missing values (if any)

So, what is it about?

Check for missing values in the dataset and fill the missing values using appropriate method. Provide justification for the method taken.

Code of Conduct

• I agree to follow this project's Code of Conduct

[Feature]: Data Cleaning: Remove Non-Discriminatory Features

So, what is it about?

You have to clean the data by removing features that do not contribute to the model’s predictive power by identifying Single-Valued Functon.

Code of Conduct

• I agree to follow this project's Code of Conduct